Review ArticleStudy QoS Optimization and Energy Saving Techniques in CloudFog Edge and IoT

Zhiguo Qu12 Yilin Wang12 Le Sun 12 Dandan Peng12 and Zheng Li12

1Engineering Research Center of Digital Forensics Ministry of Education Nanjing China2Jiangsu Collaborative Innovation Center of Atmospheric Environment and Equipment Technology (CICAEET)Nanjing University of Information Science amp Technology 210044 Nanjing China

Correspondence should be addressed to Le Sun sunle2009gmailcom

Received 6 December 2019 Revised 22 January 2020 Accepted 7 February 2020 Published 16 March 2020

Guest Editor Xuyun Zhang

Copyright copy 2020 Zhiguo Qu et al is is an open access article distributed under the Creative Commons Attribution Licensewhich permits unrestricted use distribution and reproduction in any medium provided the original work is properly cited

With an increase of service usersrsquo demands on high quality of services (QoS) more and more efficient service computing modelsare proposede development of cloud computing fog computing and edge computing brings a number of challenges eg QoSoptimization and energy saving We do a comprehensive survey on QoS optimization and energy saving in cloud computing fogcomputing edge computing and IoT environments We summarize the main challenges and analyze corresponding solutionsproposed by existing worksis survey aims to help readers have a deeper understanding on the concepts of different computingmodels and study the techniques of QoS optimization and energy saving in these models

1 Introduction

With the development of the Internet more and morecomputing techniques are developed In this situation anincreasing amount of data needs to be processed e in-crease of usersrsquo requirements causes the development ofdifferent types of computing models such as cloud com-puting fog computing and edge computing

Cloud computing is an early computing model that hasmade great contributions to data processing It providesconvenient and quick network access to shared configurableresources such as networks and servers In addition pro-visioning and publishing these resources do not requiremuch administration and interaction of service providers[1] e structure of cloud computing is shown in Figure 1

Due to the development of the IoTand the increasing needsof people the IoTsystem based on cloud computing faces somelimitations In this situation cloud computing cannot play agood role in large-scale or heterogeneous conditions [3]erefore a new computing model called fog computing isdeveloped on the basis of cloud computing Compared withcloud computing themain advantage of fog computing is that itextends cloud resources to the network edge erefore fog

computing can facilitate the management of resources andservices [4]e structure of fog computing is shown in Figure 2

Edge computing allows operations to be performed onthe edge of a network [2] Edge computing refers to all theresources of computing and network from data sources tocloud data centers In edge computing the flow of computing isbidirectional and things in edge computing can both consumedata and produce data at is they can not only ask the cloudfor services but also carries out computing jobs in the cloud [2]e structure of edge computing is shown in Figure 3

e most popular embodiment of edge computing is theMECwhich refers to the technology of performing computation-intensive and delay-sensitive tasks for mobile devices And itstheory is collecting a large amount of free computing power andstorage resources located at the edge of a networke EuropeanTelecommunication Standards Institutewas the first to define it asa computingmodelMECprovides the capabilities of informationtechnology and cloud computing at the network edge

e IoT is created by the diffusion of sensors actuatorsand other devices in the communication driven networke development of wireless technologies such as thewireless sensor network technology and actuator nodespromotes the development of the IoT technology With the

HindawiComplexityVolume 2020 Article ID 8964165 16 pageshttpsdoiorg10115520208964165

development of the IoT its application has gradually ex-panded to cover increasingly wider domains However italways aims to make computers perceive information [6]

is paper investigates the important papers related tothese computing models For each paper we point out theproblems it aims to solve and introduce the solutions it pro-posesemain contribution of this paper is as follows (1) do acomprehensive survey on the techniques of QoS optimizationand energy saving in different computing models (2) classifypapers according to the problems solved by the reviewedworks and (3) compare and summarize the main features ofeach type of paper e structure of this paper is Section 2studies five energy saving techniques under different com-puting models and Section 3 concludes this paper

2 QoS Optimization and Energy savingTechniques in Different Computing Models

In this section we introduce the main works of QoS opti-mization and energy saving techniques in different

computing models We categorize these works in terms ofthe means they use to achieve the objective of QoS opti-mization and energy saving which are (1) quality of service(QoS) guarantee or service-level agreement (SLA) assurance(2) resource management and allocation (3) scientificworkflow execution (4) server optimization and (5) loadbalancing

21 QoS Guarantee or SLA Assurance Improving QoS orreducing SLA violations can effectively guarantee thetransmission bandwidth reduce the transmission delay andreduce the packet loss rate of data Striking a balance be-tween QoS and limited resources can achieve energy saving

211 Cloud Computing Mazzucco et al [7] let cloud serviceproviders get the maximum benefit by reducing powerconsumption In addition they introduced and evaluatedthe policy of dynamic allocation of powering serversrsquoswitches It can optimize usersrsquo experience while consuming

Data

Data producer Data consumer

Request

Result

Figure 1 Structure of cloud computing [2]

Figure 2 Structure of fog computing [5]

2 Complexity

the least amount of power He et al [8] proposed a service-based system supporting keyword search in which dif-ferent search keywords represent different tasks ismethod can help unprofessional service users build ser-vice-oriented systems Sun et al [9] proposed a cloudservice selection method to measure and aggregate thenonlinear relationship between standards And a frame-work based on priority is designed to determine the criteriarelationships and weights when historical information isinsufficient Mazzucco and Dyachuk [10] were also com-mitted to making cloud service providers obtain the largestprofits ey proposed the dynamic distribution strategy ofpowering server switch e strategy not only enables usersto get good service but also reduces power consumptione number of live servers determines the state of thesystem but running or closing a server cannot be done in aflash So it is important to take into account of the timeGiven the short time required for the server switch for-mula (1) [10] represents the cost of changing the number ofservers running per unit time In order to make users havea good experience this paper further uses a forecastingmethod to accurately predict the usersrsquo time-changingneeds

Q Δnt

1113944

l

i1di + kre3

⎛⎝ ⎞⎠ (1)

where t represents the observation time Δn represents thenumber of servers whose state change over time di

represents the cost of the state change of a hardwarecomponent e3 represents the energy consumed in a unittime to change the state k represents the average time tochange the state of a server and l represents the amount ofcomponent

Mazzucco et al [7] and Mazzucco and Dyachuk [10]both explore strategies of reducing the power cost of runninga data center and changing the on-off state of the serversetwo strategies can maximize the usersrsquo experience and saveenergy at the same time eir difference is that Mazzuccoand Dyachuk [10] believes that it is impossible to accuratelypredict the changes of usersrsquo needs over time So comparedwith paper [7] the strategy proposed in paper [10] is fault-tolerant He et al [11] proposed three service selectionmethods that support QoS and can combine multitenantservice-based systems ese three methods can achievethree degrees of multitenant maturity which is more effi-cient than the traditional single-user approach Sun et al[12] proposed a unified semantic model to describe cloudservice is model expands the basic structure of unifiedservice description language And it defines a transactionmodule to model the rating system for cloud services fromvarious perspectives So it can improve the ability of themodel on service ranking In addition an annotation systemis put forward to enrich the language expression Wang et al[13] proposed a fault-tolerant strategy based on multitenantservice criticality which can provide redundancy for keycomponent services evaluating the criticality of eachcomponent service to determine the optimal fault-tolerantpolicy erefore the quality of the multitenant based ser-vice system can be guaranteed Mustafa et al [14] leveragedthe notion of workload consolidation to improve energyefficiency by putting incoming jobs on as few servers aspossible e concept of SLA is also imported to minimizethe total SLA violations Given that a change in workloadchanges the utilization of CPU required over time So anintegral function (formula (2) [14]) is used to represent thetotal energy (E) consumed by a server (S) operation

E 1113946t1

t0P(u(t))dt (2)

where P is the amount of power consumed by the server interms of CPU utilization (u) in time t

Bi et al [15] established an architecture that can ad-ministrate itself in cloud data centers firstlye architectureis suitable for web application services with several levels andhas virtualization mechanismen a mixed queuing modelis proposed to decide the number of virtual machines (VMs)in each layer of application service environments Formula(3) [15] is used to represent the local profit that can be madeby the ith virtualization application service environmentFinally a problem of misalignment restrained optimizationand a heuristic mixed optimization algorithm are proposedBoth of them can make more revenues and meet require-ments of different customers

Pi(E) Revenue(E) + Penalty(E) + Loss(E) + Cost(E)

(3)

where Revenue(E) Penalty(E) Loss(E) and Cost(E) re-spectively represent the total benefit penalty loss and costof VMs

Singh et al [16] proposed a technology named STARwhich can manage resources itself in the cloud computing

Data

Data

Data producer

Data producerconsumer

RequestResult

Edge

Computing offloadData cachingstorage

data processingRequest distribution

Service deliveryIoT management

Privacy protection

Figure 3 Structure of edge computing [2]

Complexity 3

environment and reduce SLA violations So the paymentefficiency of cloud services can be improved Beloglazov andBuyya [17] proposed a system to manage energy in the clouddata center By continuously integrating VMs and dy-namically redistributing VMs the system can achieve thegoal of saving energy and providing a high QoS level at thesame time Guazzone et al [18] proposed an automaticmanagement system (see Figure 4) for resources to providecertain QoS levels and reduce energy consumption Re-source manager of the framework in Figure 4 combinesvirtualization technologies and control-theoretic technolo-gies Virtualization technologies deploy each application toindependent VM And control-theoretic technologies realizethe automatic management of computer performance andenergy consumption In addition the resource managerconsists of several independent components named Ap-plication Manager Physical Machine Manager and Mi-gration Manager Different from traditional static methodsthis method can both fit the changing workloads dynami-cally and achieve remarkable results in reducing QoS vio-lations Sun et al [19] established a model to simplify thedecision of cloud resource allocation and realize the inde-pendent allocation of resources e optimal resourceconfiguration can be obtained so the QoS requirements canbe well met Siddesh and Srinivasa [20] explored theproblems of dynamic resource allocation and SLA assuranceey proposed a framework to deal with heterogeneousworkload types by dynamically planning computing capacityand assessing riskse framework uses scheduling methodsto reduce SLA violation risks and maximize revenues inresource allocation

Garg et al [21] proposed a resource allocation strategyfor VM dynamic allocation e strategy can improve re-source utilization increase providersrsquo profits and reduceSLA violations Jing et al [22] proposed a new dynamicallocating technique using the mixed queue model meetingcustomersrsquo different requirements of performance by pro-viding virtualized resources to each layer of virtualizedapplication services All these methods can reasonablyconfigure resources in the cloud data center improve systemperformance reduce additional costs of using resources andmeet the required QoS

Qi et al [23] proposed a QoS-aware VM schedulingstrategy named QVMS to satisfy QoS Firstly the schedulingproblem is transformed into a problem with several ob-jectives And then the optimal VM migration method isfound according to the genetic algorithm e schedulingstrategy can effectively manage resources in the networkphysical system thus reducing the energy consumption andimproving QoS levels Qi et al [25] proposed a servicerecommendation strategy by considering the time factor toimprove the traditional location-sensitive hash technologye policy emphasizes the influence of dynamic factors onQoS and the protection of user privacy

Table 1 shows a summary of the abovementioned workse solution of the problems in Table 1 can improve QoS incloud computing environment Server management refers todynamically allocating powering serversrsquo switches Work-loads consolidation refers to combining work to save energy

VM management refers to reasonable scheduling or inte-gration of VMs to achieve better performance Self-man-agement refers to the realization of self-management ofresources which can achieve higher efficiency Resourcemanagement refers to the correct allocation of resources toreduce waste Service management is about making rea-sonable service choices [26]

212 Fog Computing Gu et al [27] used fog computing toprocess a large amount of data generated by medical devicesand built Fog Computing Supported Medical Cyber-Phys-ical System (FC-MCPS) In order to reduce the cost of FC-MCPS research studies were carried out on the joint of basestation task assignment and VM layout e problem ismodeled as a mixed integer linear programming (MILP) Atwo-stage heuristic algorithm based on linear programming(LP) is proposed to solve the problem Ni et al [28] proposeda resource allocation approach based on fog computingwhich enables users to select resources independently Inaddition this approach takes into account the price and timerequired to finish the job Formula (4) [28] is used to definethe credibility BCreij of Resource Rj received from useriwhen the user interacts with Resource Ri

BCreij ω1λresp + ω2cexec + ω3 1 minus ηreboot( 1113857 + ω4μrel

(4)

where the value of ωkϵ[0 1] 11139364k1 ωk 1 which can be

determined by the user or the actual situation λresp cexecηreboot and μrel are the response speed of the corre-sponding index service the efficiency of execution the speedof restart and the reliability respectively

213 Edge Computing Wei et al [29] proposed a unifiedframework in the sustainable edge computing to save energyincluding the energy that is distributed and renewable Andthe architecture can combine the system that supply energyand edge services which can make full use of renewableenergy and provide better QoS Lai et al [30] proposed anoptimized allocation method for edge usersemethod cannot only maximize the amount of resources allocated tousers but also consider the dynamic QoS level of users Soedge user allocation problem can be made more general andimproving the quality of experience

214 MEC Xu et al [31] used block chain to improve thetraditional crowdsourcing technology Firstly they proposeda mobile crowdsourcing framework using block chaintechnology to protect user privacy en they used dynamicprogramming strategy of clustering algorithm to classifyrequesters Finally they generated service policies to balanceprofits and energy consumption

215 IoT Rolik et al [32] proposed a method to build aframework of IoT infrastructure based on microcloud emethod can help users use resources rationally reduce thecost of managing infrastructure and improve the quality of

4 Complexity

life of consumers He et al [33] proposed a dynamic networkslice strategy e network slice can be dynamically adjustedaccording to the time-varying resource demands is methodcan improve the utilization of the underlying resources andbetter meet different QoS demands Yao and Ansari [34]proposed an algorithm to determine the number of VMs to berented and to control the power supply us the cost of thesystem can be minimized and the QoS can be improvedFormula (5) [34] is used to limit the delay requirement of QoSe total delay must not exceed the computation deadline ofeach task and the total delay is composed of wireless trans-mission delay and fog processing delay

tci + t

wi leDi foralliϵN (5)

where c and w respectively represents fog processing andwireless transmission i denotes a location tc

i represents thedelay of processing tw

i represents the delay of wirelesstransmission Di denotes the deadline and N denotes dif-ferent locations

22 Resource Management and Allocation Rational alloca-tion of resources is an effective means to save energy

221 Cloud Computing Wang et al [35] introduced anallocation method based on distributed multiagent to al-locate VMs to physical machines e method can realize

VM consolidation and consider the migration costs si-multaneously In addition a VM migration mechanismbased on local negotiation is proposed to avoid unnecessaryVM migration costs Hassan et al [37] established a for-mulation of universal problem and proposed a heuristicalgorithm which has optimal parameters Under this for-mulation dynamic resource allocation can be made to meetthe QoS requirements of applications And the cost neededfor dynamic resource allocation can be minimized with thisalgorithm Wu et al [38] proposed a scheduling algorithmbased on the technology that can scale the voltage frequencydynamically in cloud computing e algorithm can allocateresources for performing tasks and realize low powerconsumption network infrastructure Compared with otherschemes this scheme not only sacrifices the performance ofexecution operations but also saves more energy

Sarbazi and Zomaya [45] used two job consolidationheuristic methods to save energy One is MaxUtil to betterutilize resources and the other is Energy-Conscious TaskConsolidation to reduce energy consumption ese twomethods can promote the concurrent execution of multipletasks and improve the energy efficiency Hsu et al [46]proposed a job consolidation technique to minimize energyconsumption Formula (6) [46] defines the energy con-sumption of VM Vi from time t0 to tm in the cluster isdefined And formula (7) [46] defines the total energyconsumption in a virtual cluster VCk in the period Inaddition the proposed technique limits the CPU usage and

Reference machine

Reference machine

SLA

SLA

Application A1

Application An

TIER 1

TIER 2

TIER 3

TIER 1

TIER 2

TIER 3

Cloud infrastructure

VM

VM

VM

VM

VM

VM

Performance measuresEnergy consumptions

Resource costs

Resourcemanager

Migrationmanager

Physicalmachinemanagers

Applicationmanagers

Figure 4 Framework of the proposed three-fold system [18]

Complexity 5

Tabl

e1

Worksummaryof

QoS

guaranteeing

orSL

Aassurancein

clou

dcompu

ting

Subp

roblem

sSolutio

nsLiteratures

Advantages

Server

managem

ent

Apo

licyof

dynamicallocatio

nof

poweringserverss

witches

[71

0]Maxim

izes

benefitsim

proves

QoSa

ndminim

izes

power

consum

ption

Workloads

consolidation

Atechniqu

eforconsolidatingworkloads

[14]

Achievesenergy

saving

sby

usingfewestserverswhile

redu

cing

SLA

violations

VM

managem

ent

Anarchitecturethat

canadministrate

itselfa

ndamixed

queuingmod

el[15]

Makes

morerevenu

esandmeets

different

requ

irem

ents

ofcustom

ersanddecidesthe

numberof

VMsforeach

layerof

avirtuala

pplication

Asystem

tointegrateanddynamically

redistribu

teVMs

[17]

Achievesthe

goalofsaving

energy

throughVM

integrationandprovidesah

ighQoS

levelat

thesametim

e

AQoS-awareVM

schedu

lingstrategy

named

QVMS

[23]

Effectiv

elymanages

resourcesin

thenetworkph

ysical

system

toredu

cetheenergy

consum

ptionandim

proves

QoS

AVM

integrationmetho

dwith

severaltargets

[24]

Savesenergy

andredu

cesSL

Aviolations

byapplying

different

strategies

todifferent

load

states

oftheho

st

Self-managem

ent

Atechno

logy

named

STAR

[16]

Redu

cesSL

Aviolations

andim

proves

paym

ente

fficiency

ofclou

dservices

Adynamic

resource

managem

entsystem

[18]

Self-manages

theresourceso

fcloud

infrastructurestoprovidea

ppropriateQoS

andfitsthe

changing

workloads

dynamically

Resource

managem

ent

Amod

elthat

canrealizetheindepend

entallocatio

nof

resources

[19]

Obtains

theop

timal

resource

confi

guratio

nmeets

theQoS

requ

irem

entsand

provides

econ

omical

clou

dresources

Adynamic

resource

allocatio

nstrategy

[20

21]

Redu

cesSL

Aandmaxim

izes

revenu

esandresource

utilizatio

non

theclou

d

Amixed

queuemod

el[22]

Reason

ably

confi

gurestheresourcesin

theclou

ddata

centerimproves

thesystem

performancereduces

theadditio

nalcosto

fusin

gresourcesmeetstherequ

ired

QoSand

provides

virtualresou

rces

toeach

layerof

virtuala

pplicationservices

Service

managem

ent

Aun

ified

semantic

mod

elthat

candescribe

clou

dservice

[12]

Improves

theability

ofmod

elon

servicerank

ingandenriches

thelang

uage

expressio

n

Arecommendatio

nservicestrategy

[25]

Emph

asizes

theinflu

ence

oftim

efactorson

QoS

andim

proves

thetradition

allocatio

n-sensitive

hash

techno

logy

toprotectusersrsquoprivacy

Aclou

dserviceselectionmetho

dusingfuzzymeasure

and

Cho

quet

integral

andafram

eworkbasedon

priority

[9]

Selectsservicewhenhistorical

inform

ationisinsufficientto

determ

inethecriteria

relatio

nships

andweigh

ts

reeserviceselectionmetho

dsthat

supp

ortQ

oSandcan

combine

multitenants

ervice-based

system

s[11]

Achievesthree

degreeso

fmultitenantm

aturity

which

ismoreeffi

cientthanthetradition

alsin

gle-user

approach

Afault-tolerant

strategy

basedon

multitenants

ervice

criticality

[13]

Guaranteesthequ

ality

ofthemultitenant-basedservicesystem

Aservice-basedsystem

supp

ortin

gkeyw

ordsearch

[8]

Effectiv

elyhelpssystem

engineerswho

areno

tfam

iliar

with

service-oriented

architecture

techno

logy

tobu

ildservice-oriented

system

s

6 Complexity

merges tasks in virtual clusters Once a task migrationhappens the energy cost model will take into account thenetwork latency Sarbazi-Azad and Zomaya [45] and Hsuet al [46] both maximize the benefit of cloud resources byusing task merging techniques Sarbazi-Azad and Zomaya[45] uses a greedy algorithm called MaxUtil While Hsuet al [46] takes into account the network latency associatedwith task migration So in [46] a 17 improvement isachieved over MaxUtil

E0m Vi( 1113857 1113944m

t0Et Vi( 1113857 (6)

E0m VCk( 1113857 1113944n

i0E0m Vi( 1113857 (7)

where Et is the energy consumption in unit time and n is thenumber of VMs in the cluster

Hsu et al [47] proposed a task integration technologybased on the energy perception According to the charac-teristics of most cloud systems the principle of using 70CPU is proposed to administrate job integration amongvirtual clusters is technology is very effective in reducingthe amount of energy consumed in cloud systems bymerging tasks Panda and Jana [48] proposed an algorithmwith several criteria to combine tasks e algorithm notonly considers the time needed for processing jobs but alsoconsiders the utilization rate of VMs And the algorithm ismore energy efficient because it takes into account not onlythe processing time but also the utilization rate of VMsWang and Su [39] proposed a resource allocation algorithmto deal with wide range of communication between nodes incloud environment is algorithm uses recognition tech-nology to dynamically distribute jobs and nodes accordingto computing ability and factors of storage And it can re-duce the traffic when allocating resources because it usesdynamic hierarchy Lin et al [40] proposed a dynamicauction approach for resource allocation e approach canensure that even if there are many users and resourcesproviders will have reasonable profits and computing re-sources will be allocated correctly Yazir et al [41] proposeda new method to manage resources dynamically and au-tonomously Firstly resource management is split into jobsand each job is executed by autonomous nodes Secondautonomous nodes use the method called PROMETHEE toconfigure resources Krishnajyothi [36] proposed a frame-work which can implement parallel task processing to solvethe problem of low efficiency when submitting large tasksCompared with the static framework this framework candynamically allocate VMs thus reducing costs and the timeof processing tasks Lin et al [42] proposed a method toallocate resources dynamically by using thresholds Becausethis method uses the threshold value it can optimize thereallocation of resources improve the usage of resourcesand reduce the cost Xu et al [43] proposed a data placementstrategy named IDP for the data generated by IoTs devices toachieve reasonable data placement In this way the privacyof these data can be protected while resources are allocatedreasonably Jo et al [44] proposed a computing offload

framework under 5G network e framework transfers thecomputing burden to the cloud thus reducing the com-puting load of clients and the communication cost

Table 2 shows a summary of the abovementioned workse problem of resource allocation and management incloud computing can be divided into problems in Table 2VM management is about a reasonable configuration ofVMs Resource allocation represents the dynamic andflexible allocation of resources Task integration refers tocombining tasks to save energy and improve efficiency

222 Fog Computing Yin et al [49] established a newmodel of scheduling jobs which applies containers In orderto make sure that jobs can be finished on time a jobscheduling algorithm is developed e algorithm can alsooptimize the number of tasks that can be performed togetheron the nodes in fog computing And this paper proposes aredistribution mechanism to shorten the delay of tasksese methods are very effective in reducing task delaysAazam and Huh [50] established a framework to admin-istrate resources effectively in the mode of fog computingConsidering that there are various types of objects anddevices the connection between them may be volatile So amethod for predicting and administrating resources isproposed e method considers that any object or devicecan quit using resources at anytime Cuong et al [5] studiedthe allocating resources jointly problem and the problem ofcarbon footprint minimization in fog data center Formula(8) [5] is used to denote the energy consumption of serversIn addition a distributed algorithm is proposed to solve theproblem of wide range optimization

P(y) C middot Pidle + Ppeak minus Pidle1113872 1113873 middot y middot κ (8)

where P(y) represents the power supply required by theservers in a data center y represents the video stream κdenotes a conversion factor that converts the video streaminto workload C represents the data centerrsquos load capacityand Pidle and Ppeak respectively represent the idle power andpeak power of the servers

Jia et al [51] studied the problem of computing resourceallocation in fog computing network with three levelsFirstly the problem of resource allocation is transformedinto a bilateral matching optimal problem And then a bi-matching approach is proposed for this problem which canimprove the performance of the system and obtain highercost efficiency Zhang et al [52] proposed a framework forjoint optimization under fog computing to allocate fognodesrsquo finite computing resources e framework canachieve the best allocation and effectively improve thenetworksrsquo performance Tan et al [53] presented a method toallocate computing and communication resources emethod transfers computing jobs to remote cloud and nodesand simplifies edge nodesrsquo computing and computing en-ergy Vasconcelos et al [54] developed a platform to allocateresources accessible to client devices in fog computingenvironment allocating the resources of devices near thehost to meet the applications needs for rapid response tocomputing resources Aazam et al [55] presented a method

Complexity 7

to estimate and manage resource in fog computing emethod is based on the fluctuation of customer abandon-ment probability type and price of service and so on

Table 3 shows a summary of the abovementioned workse problems in Table 3 are also derived from resourceallocation and management problems Task allocationrepresents the scheduling and redistribution of tasks Re-source allocation is still about the dynamic and flexibleallocation of resources Low latency refers to taking shorttime to configure and manage resources which can improveefficiency

223 Edge Computing Tung et al [56] proposed a newframework for resource allocation based on market needse resources come from edge nodes (ENs) with limitedheterogeneous capabilities and are allocated to multiplecompeting services on the network edge Generating amarket equilibrium solution by reasonably pricing ENs canobtain the maximum utilization of marginal computingresources Xu et al [57] proposed a strategy to optimizeoffloading and privacy protection is strategy shifts tasksfirstly to improve the resource utilization of resource-limited

edge cells And then it balances QoS performance andprivacy protection to achieve joint optimization

Xu et al [58] proposed an offload strategy for edgecomputing under 5G network which uses block chaintechnologye optimal strategy is further obtained by usingthe balanced offloading method It solves the problem ofdata loss under the condition of transmission delay which iscaused by the uneven requirements of user equipments onresources Xu et al [59] proposed a computational off-loading method named EACO to reduce the energy con-sumption in smart computing models Figure 5 showsarchitecture of smart edge computing where the shortestpath is used to unload tasks EACO uses genetic algorithmsto reduce the energy consumption for operating edgecomputing nodes and improve the efficiency of performingcomplex computing tasks Xu et al [60] proposed a com-putational offloading strategy for edge computing to protectthe privacy of interconnected vehicle networks ey firstlyanalyzed privacy conflicts of tasks And then they designedthe communication route to obtain routing vehicles whichcan achieve the optimization of several objectives Yetinget al [61] proposed a unique resource allocation mechanisme mechanism takes each individual task as the basis for

Table 2 Work summary of resource allocation and management in cloud computing

Subproblems Solutions Literatures Advantages

VMmanagement

A VM allocation method based on severaldistributed agents [35] Centralizes VMs to physical machines and reduces the

overall energy costA framework which can implement parallel

task processing [36] Dynamically allocates VMs for large tasks

Resourceallocation

A general problem formula and a heuristicalgorithm with optimization parameters [37]

Dynamically allocates resources according to QoSrequirements and realizes energy saving by optimizing the

number of serversA scheduling algorithm based on thetechnology that can scale the voltage

frequency dynamically[38] Ensures the performance of executing jobs while

implementing green computing

A fuzzy pattern recognition technology [39] Reduces the traffic when allocating resources and usesdynamic hierarchy

A dynamic auction approach [40]Guarantees profits of providers and allocates resourcescorrectly when there are a large amount of users and

resources

PROMETHEE [41]Allows node agents to decompose and execute tasksautonomously to improve the flexibility of resource

allocation

A method using thresholds [42]Optimizes resource reallocation improves resource usagereduces cost and studies resource allocation strategies at the

application level

IoT-oriented data placement (IDP) method [43] Protects data privacy allocates resources reasonably andfocuses on the placement method of IoT data

A computing offload framework under 5Gnetwork [44] Reduces the computing load of clients and the

communication cost

Taskconsolidation

Two job consolidation heuristics namedldquoMaxUtilrdquo and energy-conscious task

consolidation[45] Promotes concurrent execution of multiple tasks and

improves energy efficiency

A job consolidation technique aiming atenergy saving [46 47]

Reduces the power consumption of cloud system protectsdata privacy allocates resources reasonably limits CPU

usage and merges tasks in virtual clusters

An algorithm with several criteria [48]Takes into account the job processing time and the VM

utilization rate simultaneously Dramatically reduces energyconsumption compared with state-of-the-art works

8 Complexity

resource allocation rather than for the whole service It reducesthe packet loss rate and saves energy by unloading services

224 MEC Chen et al [62] studied the problem of com-puting unloading with several users in the environment ofMEC with wireless interference which have many channels Inaddition a distributed algorithm for computing unloading isdevelopede algorithm can perform the unloading well evenwhen there are a large number of users Gao et al [63] built aquadratic binary program which is able to assign tasks inmobile cloud computing environment Two algorithms arepresented to obtain the optimal solution Both of these heuristicalgorithms can effectively solve the task assignment problemXu et al [64] proposed an offloadingmethod using block chaintechnology It can guarantee the loss of data in offloading tasksunder edge computing And it can solve the problem of re-source requests out of proportion due to the limited load of

edge computing equipment during task transfer Yifei et al [65]proposed a model-free reinforcement learning framework tosolve the problem of computational unloadingis model canbe applied to the computational unloading with time-changingcomputing requests

225 IoT Barcelo et al [66] expressed the problem ofservice allocation [67] as a mixed flow problem with min-imum cost which can be solved by LP solving this serviceallocation problem can solve the problems of unbalancednetwork load and delay of end-to-end service And it canalso figure out the problem of excessive consumption ofelectricity brought by the architecture of centralized cloudAngelakis et al [68] assigned the requirements of servicesresources to heterogeneous network interfaces of equipmentsSo more heterogeneous network interfaces can be used by alarge amount of services

Mobile devices

Mobile devices ECN

ECN

ECNECN

AP

AP

AP

APAP

Figure 5 An architecture of smart edge computing [59]

Table 3 Work summary of resource allocation and management in fog computing

Problems Solutions Literatures AdvantagesTaskallocation

A new model of scheduling jobs and aredistribution mechanism [49] Finishes jobs on time optimizes the number of tasks and

shortens the delay of tasks

Resourceallocation

A framework to administrate resources and amethod to predict and administrate resources [50]

Assists service providers to predict the amount of availableresources based on different types of service customers anddeals with the phenomenon that objects or devices withdraw

from resource utilization at any time

A bi-matching approach [51] Improves the performance of the system and obtains highercost efficiency

A framework for joint optimization [52] Optimizes resource allocation and improves networkperformance

A method to allocate computing andcommunication resources [53]

Allocates computing and communication resources transferscomputing jobs to remote cloud and nodes simplifies edge

nodesrsquo computing and saves computing energyA platform to allocate resources accessible to

client devices [54] Enables rapid response to computing resources and allocatesthe resources of devices near the host

A framework to manage resources [55] Considers the fluctuation of the customer abandonmentprobability and service types

Low latency A distributed algorithm [5] Solves the problem of wide range optimization and allocatesresources jointly

Complexity 9

Li et al [69] proposed communication framework in 5Gand studied the problem of allocating power andchannels So the signal data in the channel can beavailable and the total energy efficiency can be maximumFormula (9) [69] shows how to calculate the energy ef-ficiency of a system

U 1113944K

k11113944

M

i1EES

iK + 1113944K

K11113944

N

i1EEA

jk (9)

where EESiK and EEA

jk respectively denote the energy ef-ficiency of sensor S and actuator A on channels e sets ofsensors actuators and channels are respectively repre-sented as s S1 S2 SM1113864 1113865 A A1 A2 AN1113864 1113865 andC C1 C2 CK1113864 1113865

Liu et al [70] studied the problem of allocating resourcesefficiently on IoT that supplies wireless power In thismethod users are first grouped into accessible channelsAnd then power distribution of users grouped in the samechannel is studied to improve throughput of the networkis method can allocate finite resources to a large group ofusers Ejaz and Ibnkahla [71] proposed the resource allo-cation framework with several bands under cognitive 5GIoT In the highly dynamic environment of the IoTmultiband method can manage resources more flexiblyand reduce more energy consumption In addition areconstruction approach with several levels is proposed toallocate resources reasonably for applications with dif-ferent needs of QoS Colistra et al [72] proposed aprotocol which is distributed and optimal to allocateresources in heterogeneous IoT Because this protocol hasexcellent adaptability when changing topology of net-work it can distribute resources evenly among nodes Jianet al [73] proposed a multilevel allocating resources al-gorithm for IoT communication using advanced tech-nology e algorithm uses hierarchical structure and hasfast data processing rate and very low latency in a satu-rated or not saturated environment

Zheng and Liu [74] proposed a new algorithm to allocatebandwidth dynamically for controlling remote computers inthe IoT is method can reduce the error of signal re-construction under the same bandwidth and make thebandwidth allocation of IoT more reasonable Gai and Qiu[75] used reinforcement learning mechanisms to allocateresources to achieve high Quality of Experience ismethod can effectively solve the resource allocation prob-lems caused by the mismatch of service quality and complexservice providing condition in the IoT

Table 4 shows a summary of the above works eproblem in Table 4 represents the realization of dynamic andflexible allocation of resources e resources here canrepresent channels bandwidth and power

23 Scientific Workflow Execution Implementing scientificworkflows especially in heterogeneous environments canreduce resource waste and reduce energy costs Scientificworkflow can be obtained by reasonably allocating resourcesand dynamically deploying VMs

231 Cloud Computing Xu et al [76] proposed a resourceallocation method called EnRealan to solve the problem ofenergy consumption e dynamic deployment of VMs isgenerally adopted to execute scientific workflows Bousselmiet al [77] proposed a scheduling method based on energyperception for executing scientific workflows in cloudcomputing At first an algorithm of splitting workflows forenergy minimization is presented which can achieve a highparallelism without huge energy consumption en aheuristic algorithm used to optimize cat swarm is proposedfor the created partitions e algorithm can minimize thetotal consumption of energy and the execution time ofworkflows Sonia et al [78] proposed a workflow schedulingmethod with several objects and hybrid particle swarmoptimization algorithm In addition a method for dy-namically scaling voltage and frequency is proposed emethod canmake the processors work at any voltage level soas to minimize the energy consumption in the process ofworkflow scheduling Both Bousselmi et al [77] and Soniaet al [78] use scheduling method to achieve scientificworkflows and study the problem of energy consumptione difference is that Bousselmi et al [77] focuses on in-tensive computing tasks while Sonia et al [78] focuses onworkflow scheduling on heterogeneous computing systems

Cao [79] established a scheduling algorithm of scientificworkflows with an objective of energy savingis algorithmcan enable service providers to gain high profits and reduceusersrsquo overhead at the same time Li et al [80] proposed ascheduling algorithm based on cloud computing which canminimize cost of performing workflows within a specifiedtime In addition the rented VM was modified to save costfurther Khaleel and Zhu [81] proposed a scheduling algo-rithm and took scientific workflows as a model to make fulluse of cloud resources and save energy Shi et al [82]designed a flexible resource allocation and job schedulingmechanism to implement scientific workflows Because thismechanism can implement scientific workflows withinprescribed budgets and deadlines so it can work better thanother mechanisms

Table 5 shows a summary of the abovementioned workse problems in Table 5 are derived from the imple-mentation of the scientific workflow VM deployment refersto the rational allocation of VMs Workflow scheduling refersto reducing the scheduling energy and time In addition itrefers to the scheduling of the workflow on heterogeneoussystems Cost reduction refers to reducing the cost of workflowexecution Effective implementation is about scientific work-flow execution within a specified budget and time

24 Server Optimization Server optimization is also a goodway to save energy e goal of optimizing the server can beachieved by uninstalling unnecessary servers or consoli-dating servers as well as by reasonably scheduling tasksUnlike QoS optimization server optimization aims to op-timize the number of used servers improves the energyefficiency of servers and consolidates servers However QoSoptimization studies how tomake users get better experienceand meet their needs

10 Complexity

241 Cloud Computing Ge et al [83] proposed a game-theoretic method and transformed the problem of mini-mizing energy into a congestion game All mobile devices inthis method are participants in the game e methodchooses a server to unload the computation tasks to optimizeQoS levels and save energy which can optimize the systemand save energy Wang et al [84] proposed a MapReduce-based multitask scheduling algorithm to achieve the ob-jective of energy saving is model is a two-layer modelwhich considers the impact of server performance changeson energy consumption and the limitation of networkbandwidth In addition a local search operator is designedbased on which a two-layer genetic algorithm is proposede algorithm can schedule tens of thousands of tasks incloud and achieve large-scale optimization Yanggratoke

et al [85] proposed a general generic gossip protocol aimingat allocating resources in cloud environment An instanti-ation of this protocol was developed to enable server con-solidation to allocate resources to more servers to meetchanging load patterns

25 LoadBalancing Load balancing can help save energy bymanaging the number of servers and allocating resources

251 Cloud Computing Paya and Marinescu [86] intro-duced an operation model that balances cloud computingload and expands applications to save energy e principleof this model is to define an operating system e systemshould make servers run in the system as many as possible

Table 5 Work summary of scientific workflow execution in cloud computing

Problems Solutions Literatures Advantages

VM deployment A resource allocation method namedEnRealan [76] Performs scientific workflows based on energy perception

across cloud platforms

Workflowscheduling

A scheduling method based on energyperception [77]

Achieves a high parallelism without huge energyconsumption and minimizes the total consumption of

energy and execution time of workflows

A workflow scheduling method with severalobjects and hybrid particle swarm

optimization algorithm[78]

Makes the processors work at any voltage level minimizesthe energy consumption in the process of workflowscheduling and studies the scheduling problem of

workflows on heterogeneous systemsA scheduling algorithm based on various

applications [79] Enables service providers to gain high profits and reducesuser overhead at the same time

Cost reduction A scheduling algorithm based on energyperception [80 81] Minimizes the cost of performing workflows while

meeting the time constraintEffectiveimplementation

A flexible resource allocation and jobscheduling mechanism [82] Implements scientific workflows within prescribed

budgets and deadlines

Table 4 Work summary of resource allocation and management in IoT

Problems Solutions Literatures Advantages

Resourcemanagement

A distributed cloud network framework [66]Replaces the centralized architecture with a distributed cloud

architecture solves the defects of the centralized cloudarchitecture and brings people better experience

A MILP model [68] Assigns the requirement of servicesrsquo resource to heterogeneousnetwork equipment interface

A framework for communication used in5G [69]

Transforms the resource allocation problem into a power andchannel allocation problem minimizes the total energy

consumption and improves QoS levelsA low complexity channel allocation

algorithm [70] Improves throughput of the network and allocates finiteresources to a large group of users

A resource allocation framework withseveral bands under cognitive 5G IoT [71] Manages resources more flexibly and reduces energy

consumption than common single-band approach

A protocol which is distributed andoptimal to allocate resources [72]

Has excellent adaptability when changing topology of networkand dynamically manages resources in the heterogeneous IoT

environmentA multilevel allocating resourcesalgorithm for IoT communication [73] Has fast data processing rate and very low latency in both

saturated and nonsaturated environment

A new algorithm to allocate bandwidthdynamically [74]

Reduces the error of signal reconstruction under the samebandwidth and makes the bandwidth allocation of IoT more

reasonable

A reinforcement learning mechanism [75]Effectively solves the resource allocation problems caused by themismatch of service quality and complex service providing

condition in the IoT

Complexity 11

When no tasks are being performed the system should adjustservers to sleep thus energy consumption can be reducedJustafort et al [87] mainly studied the problem of workloaddistribution across cloud computing environment and pro-posed a method to solve the problem of the VM layout So thefootprint of carbon can be effectively reduced Panwar andMallick [88] proposed an algorithm to dynamically manage theload and effectively distribute the total incoming requestsbetween VMs rough efficient and uniform utilization ofresources this algorithm can achieve uniform distribution ofload between servers Yang et al [89] proposed a powermanagement mechanism to balance the load e system canmonitor VMs and dynamically allocate the resources Yanget al [90] proposed an optimization system to better allocateresources dynamically which can balance the load of VMsrunning on multiple physical machines Under this systemVMs can be migrated automatically to adjust high and lowloads without interrupting services Yang et al [89 90] manageVMs to achieve load balancing ey allocate resources dy-namically to migrate VMs which can balance workloads ondifferent physical machines e difference is that Yanget al[89] integrate a dynamic resource allocation approachwithOpenNebulaWhile Yang et al [90] focuse on avoiding serviceoutages during VM migration

Table 6 shows a summary of the abovementioned workse problems in Table 6 are from the load balancing problemServer management is about the control of the number ofservers running in the system Workload management is therational allocation of workload or tasks VMmanagement refersto configuringVMresources andmigratingVMs to adjust loads

252 Fog Computing Xu et al [91] proposed a methodcalled ldquoDRAMrdquo to dynamically allocate resources in fogcomputing environment which can avoid both too highand too low loads e method first analyzes the loadbalance of different kinds of computing nodes And then itdesigns a fog resource allocation method to achieve loadbalance which allocates resources statically and migratesservices dynamically Oueis et al [92] studied the loadbalance problem in fog computing A custom fog clus-tering algorithm is proposed to solve the problem In theproblem several users need to offload computations andall of their demands need to be handled by local com-puting cluster

253 IoT Wang et al [93] established architecture of theenergy saving targeted system in industrial IoT In additionin order to predict sleep intervals they developed a sleepscheduling and a wake protocol which provide a better wayfor energy saving

3 Conclusion

is paper did a comprehensive study of QoS optimi-zation and energy saving in cloud computing edgecomputing fog computing and IoT models We sum-marized five main problems and analyzed their solutionsproposed by existing works By conducting this surveywe aim to help readers have a deeper understanding onthe concepts of different computing models and thetechniques of QoS optimization and energy saving inthese models

e investigated papers focus on issues about ensuringQoS and reducing SLA violations and resource manage-ment In the case of QoS assurance and SLA violationsreduction the main solution of QoS assurance is efficientVM management is solution can meet customersrsquo re-quirements through reasonable scheduling and integrationof VMs Most of resource management techniques are re-alized by reasonable scheduling of resources which canreduce the waste of VMs servers and traffic

Disclosure

is manuscript is an extension of A Survey of QoS Opti-mization and Energy Saving in Cloud Edge and IoT in e9th EAI International Conference on Cloud Computing

Conflicts of Interest

e authors declare that there are no conflicts of interestregarding the publication of this paper

Acknowledgments

is work was supported by the National Natural ScienceFoundation of China (Grant no 61702274) Natural ScienceFoundation of Jiangsu Province (Grant no BK20170958)and PAPD

Table 6 Work summary of load balancing in cloud computing

Problems Solutions Literatures AdvantagesServermanagement

An operation model that can balance cloudcomputing load and expand application [86] Saves energy by managing the number of servers running in

the system

Workloadmanagement

A hybrid approach [87] Reduces the footprint of carbon and allocates workload acrosscloud computing

An algorithm to dynamically manage theload [88] Manages the load evens the load distribution between

servers and allocates tasks between VMs

VMmanagement

A green power administration mechanism [89] Monitors and jointly allocates VM resources

An optimization system [90]Migrates VMs to adjust high and low loads without

interrupting services and balances the load of VMs runningon multiple physical machines

12 Complexity

References

[1] Q Zhang L Cheng and R Boutaba ldquoCloud computingstate-of-the-art and research challengesrdquo Journal of InternetServices and Applications vol 1 no 1 pp 7ndash18 2010

[2] W Shi J Cao Q Zhang Y Li and L Xu ldquoEdge computingvision and challengesrdquo IEEE Internet of Aings Journal vol 3no 5 pp 637ndash646 2016

[3] M Iorga L Feldman R Barton M J Martin N S Gorenand C Mahmoudi ldquoFog computing conceptual modelrdquo TechRep Recommendations of the National Institute of Standardsand Technology Gaithersburg MD USA 2018

[4] Nebbiolo ldquoFog vs edge computingrdquo Tech Rep NebbioloTechnologies Inc Milpitas CA USA 2018

[5] C T Do N H Tran C Pham M G R Alam J H Son andC S Hong ldquoA proximal algorithm for joint resource allo-cation andminimizing carbon footprint in geo-distributed fogcomputingrdquo in Proceedings of the 2015 International Con-ference on Information Networking (ICOIN) pp 324ndash329IEEE Siem Reap Cambodia January 2015

[6] S Vashi J Ram J Modi S Verma and C Prakash ldquoInternetof things (IoT) a vision architectural elements and securityissuesrdquo in Proceedings of the 2017 International Conference onI-SMAC (IoT in Social Mobile Analytics and Cloud)(I-SMAC) pp 492ndash496 IEEE Coimbatore India February2017

[7] M Mazzucco D Dyachuk and R Deters ldquoMaximizing cloudprovidersrsquo revenues via energy aware allocation policiesrdquo inProceedings of the 2010 IEEE 3rd International Conference onCloud Computing pp 131ndash138 IEEE Miami FL USA July2010

[8] Q He R Zhou X Zhang et al ldquoKeyword search for buildingservice-based systemsrdquo IEEE Transactions on Software En-gineering vol 43 no 7 pp 658ndash674 2016

[9] L Sun H Dong O K Hussain F K Hussain and A X LiuldquoA framework of cloud service selection with criteria inter-actionsrdquo Future Generation Computer Systems vol 94pp 749ndash764 2019

[10] M Mazzucco and D Dyachuk ldquoOptimizing cloud providersrevenues via energy efficient server allocationrdquo SustainableComputing Informatics and Systems vol 2 no 1 pp 1ndash122012

[11] Q He J Han F Chen et al ldquoQos-aware service selection forcustomisable multi-tenant service-based systems maturityand approachesrdquo in Proceedings of the 2015 IEEE 8th Inter-national Conference on Cloud Computing pp 237ndash244 IEEENew York NY USA July 2015

[12] L Sun J Ma H Wang Y Zhang and J Yong ldquoCloud servicedescription model an extension of usdl for cloud servicesrdquoIEEE Transactions on Services Computing vol 11 no 2pp 354ndash368 2015

[13] Y Wang Q He D Ye and Y Yang ldquoFormulating criticality-based cost-effective fault tolerance strategies for multi-tenantservice-based systemsrdquo IEEE Transactions on Software En-gineering vol 44 no 3 pp 291ndash307 2017

[14] S Mustafa K Bilal S U R Malik and S A Madani ldquoSla-aware energy efficient resource management for cloud en-vironmentsrdquo IEEE Access vol 6 pp 15004ndash15020 2018

[15] J Bi H Yuan M Tie and W Tan ldquoSla-based optimisation ofvirtualised resource for multi-tier web applications in clouddata centresrdquo Enterprise Information Systems vol 9 no 7pp 743ndash767 2015

[16] S Singh I Chana and R Buyya ldquoStar sla-aware autonomicmanagement of cloud resourcesrdquo IEEE Transactions on CloudComputing p 1 2017

[17] A Beloglazov and R Buyya ldquoEnergy efficient resourcemanagement in virtualized cloud data centersrdquo in Proceedingsof the 2010 10th IEEEACM International Conference onCluster Cloud and Grid Computing pp 826ndash831 IEEEComputer Society Melbourne Australia May 2010

[18] M Guazzone C Anglano and M Canonico ldquoEnergy-effi-cient resource management for cloud computing infra-structuresrdquo in Proceedings of the 2011 IEEE AirdInternational Conference on Cloud Computing Technology andScience pp 424ndash431 IEEE Athens Greece November 2011

[19] Y Sun J White and S Eade ldquoA model-based system toautomate cloud resource allocation and optimizationrdquo inProceedings of the International Conference on Model DrivenEngineering Languages and Systems pp 18ndash34 SpringerValencia Spain October 2014

[20] G Siddesh and K Srinivasa ldquoSla-driven dynamic resourceallocation on cloudsrdquo in Proceedings of the InternationalConference on Advanced Computing Networking and Securitypp 9ndash18 Springer Surathkal India December 2011

[21] S K Garg S K Gopalaiyengar and R Buyya ldquoSla-basedresource provisioning for heterogeneous workloads in avirtualized cloud datacenterrdquo in Proceedings of the Interna-tional Conference on Algorithms and Architectures for ParallelProcessing pp 371ndash384 Springer Melbourne AustraliaOctober 2011

[22] J Bi Z Zhu and H Yuan ldquoSla-aware dynamic resourceprovisioning for profit maximization in shared cloud datacentersrdquo in Proceedings of the International Conference onHigh Performance Networking Computing and Communi-cation Systems pp 366ndash372 Springer Singapore May 2011

[23] L Qi Y Chen Y Yuan S Fu X Zhang and X Xu ldquoA Qos-aware virtual machine scheduling method for energy con-servation in cloud-based cyber-physical systemsrdquo WorldWide Web vol 4 no 3 pp 1ndash23 2019

[24] A Beloglazov J Abawajy and R Buyya ldquoEnergy-aware re-source allocation heuristics for efficient management of datacenters for cloud computingrdquo Future Generation ComputerSystems vol 28 no 5 pp 755ndash768 2012

[25] L Qi R Wang C Hu S Li Q He and X Xu ldquoTime-awaredistributed service recommendation with privacy-preserva-tionrdquo Information Sciences vol 480 pp 354ndash364 2019

[26] Z Zhai B Cheng Y Tian J Chen L Zhao and M Niu ldquoAdata-driven service creation approach for end-usersrdquo IEEEAccess vol 4 pp 9923ndash9940 2016

[27] L Gu D Zeng S Guo A Barnawi and Y Xiang ldquoCostefficient resource management in fog computing supportedmedical cyber-physical systemrdquo IEEE Transactions onEmerging Topics in Computing vol 5 no 1 pp 108ndash119 2015

[28] L Ni J Zhang C Jiang C Yan and K Yu ldquoResource al-location strategy in fog computing based on priced timed petrinetsrdquo Ieee Internet of Aings Journal vol 4 no 5pp 1216ndash1228 2017

[29] L Wei T Yang F C Delicato et al ldquoOn enabling sustainableedge computing with renewable energy resourcesrdquo IEEECommunications Magazine vol 56 no 5 pp 94ndash101 2018

[30] P Lai Q He G Cui et al ldquoEdge user allocation with dynamicquality of servicerdquo in Proceedings of the International Con-ference on Service-Oriented Computing pp 86ndash101 SpringerToulouse France October 2019

[31] X Xu Q Liu X Zhang J Zhang L Qi and W Dou ldquoAblockchain-powered crowdsourcing method with privacy

Complexity 13

preservation in mobile environmentrdquo IEEE Transactions onComputational Social Systems vol 6 no 6 pp 1407ndash1419 2019

[32] O Rolik E Zharikov and S Telenyk ldquoMicrocloud-basedarchitecture of management system for IoT infrastructuresrdquoin Proceedings of the 2016 Aird International Scientific-Practical Conference Problems of Infocommunications Scienceand Technology (PIC SampT) pp 149ndash151 IEEE KharkivUkaraine October 2016

[33] W He S Guo Y Liang R Ma X Qiu and L Shi ldquoQos-awareand resource-efficient dynamic slicingmechanism for internetof thingsrdquo Computers Materials amp Continua vol 61 no 3pp 1345ndash1364 2019

[34] J Yao and N Ansari ldquoQos-aware fog resource provisioningand mobile device power control in IoT networksrdquo IEEETransactions on Network and Service Management vol 16no 1 pp 167ndash175 2018

[35] W Wang Y Jiang and W Wu ldquoMultiagent-based resourceallocation for energy minimization in cloud computing sys-temsrdquo IEEE Transactions on Systems Man and CyberneticsSystems vol 47 no 2 pp 1ndash16 2016

[36] K Krishnajyothi ldquoParallel data processing for effective dy-namic resource allocation in the cloudrdquo International Journalof Computer Applications vol 70 no 22 pp 1ndash4 2013

[37] M M Hassan B Song M S Hossain and A Alamri ldquoEf-ficient resource scheduling for big data processing in cloudplatformrdquo in Proceedings of the International Conference onInternet and Distributed Computing Systems pp 51ndash63Springer Calabria Italy September 2014

[38] C-M Wu R-S Chang and H-Y Chan ldquoA green energy-efficient scheduling algorithm using the dvfs technique forcloud datacentersrdquo Future Generation Computer Systemsvol 37 no 7 pp 141ndash147 2014

[39] Z Wang and X Su ldquoDynamically hierarchical resource-al-location algorithm in cloud computing environmentrdquo AeJournal of Supercomputing vol 71 no 7 pp 2748ndash2766 2015

[40] W-Y Lin G-Y Lin and H-Y Wei ldquoDynamic auctionmechanism for cloud resource allocationrdquo in Proceedings ofthe 2010 10th IEEEACM International Conference on ClusterCloud and Grid Computing pp 591-592 IEEE ComputerSociety Victoria Australia May 2010

[41] Y O Yazir C Matthews R Farahbod et al ldquoDynamic re-source allocation in computing clouds using distributedmultiple criteria decision analysisrdquo in Proceedings of the 2010IEEE 3rd International Conference on Cloud Computingpp 91ndash98 IEEE Washington DC USA July 2010

[42] W Lin J Z Wang C Liang and D Qi ldquoA threshold-baseddynamic resource allocation scheme for cloud computingrdquoProcedia Engineering vol 23 no 5 pp 695ndash703 2011

[43] X Xu S Fu L Qi et al ldquoAn IoT-oriented data placementmethod with privacy preservation in cloud environmentrdquoJournal of Network and Computer Applications vol 124pp 148ndash157 2018

[44] J Bokyun P Md Jalil L Daeho and S Doug Young ldquoEf-ficient computation offloading in mobile cloud computing forvideo streaming over 5 grdquo Computers Materials amp Continuavol 61 no 2 pp 439ndash463 2019

[45] H Sarbazi-Azad and A Y Zomaya ldquoEnergy-efficient resourceutilization in cloud computingrdquo in Large Scale Network-Centric Distributed Systems pp 377ndash408 Wiley-IEEE PressHoboken NJ USA 2014

[46] C-H Hsu K D Slagter S-C Chen and Y-C ChungldquoOptimizing energy consumption with task consolidation incloudsrdquo Information Sciences vol 258 no 3 pp 452ndash4622014

[47] C-H Hsu S-C Chen C-C Lee et al ldquoEnergy-aware taskconsolidation technique for cloud computingrdquo in Proceedingsof the 2011 IEEE Aird International Conference on CloudComputing Technology and Science pp 115ndash121 IEEE Ath-ens Greece 2011

[48] S K Panda and P K Jana ldquoAn efficient task consolidationalgorithm for cloud computing systemsrdquo in Proceedings of theInternational Conference on Distributed Computing and In-ternet Technology pp 61ndash74 Springer Bhubaneswar IndiaJanuary 2016

[49] L Yin J Luo and H Luo ldquoTasks scheduling and resourceallocation in fog computing based on containers for smartmanufacturingrdquo IEEE Transactions on Industrial Informaticsvol 14 no 10 pp 4712ndash4721 2018

[50] M Aazam and E-N Huh ldquoDynamic resource provisioningthrough fog micro datacenterrdquo in Proceedings of the 2015IEEE International Conference on Pervasive Computing andCommunication Workshops (PerCom Workshops) pp 105ndash110 IEEE St Louis MO USA March 2015

[51] B Jia H Hu Y Zeng T Xu and Y Yang ldquoDouble-matchingresource allocation strategy in fog computing networks basedon cost efficiencyrdquo Journal of Communications and Networksvol 20 no 3 pp 237ndash246 2018

[52] H Zhang Y Xiao S Bu D Niyato F R Yu and Z HanldquoComputing resource allocation in three-tier IoT fog net-works a joint optimization approach combining stackelberggame and matchingrdquo IEEE Internet of Aings Journal vol 4no 5 pp 1204ndash1215 2017

[53] J Tan T-H Chang and T Q Quelc ldquoMinimum energyresource allocation in fog radio access network with fronthauland latency constraintsrdquo in Proceedings of the 2018 IEEE 19thInternational Workshop on Signal Processing Advances inWireless Communications (SPAWC) pp 1ndash5 IEEE KalamataGreece June 2018

[54] D R de Vasconcelos R M de Castro Andrade andJ N de Souza ldquoSmart shadowndashan autonomous availabilitycomputation resource allocation platform for internet ofthings in the fog computing environmentrdquo in Proceedings of the2015 International Conference on Distributed Computing inSensor Systems pp 216-217 IEEE Fortaleza Brazil June 2015

[55] M Aazam M St-Hilaire C-H Lung and I LambadarisldquoPre-fog IoT trace based probabilistic resource estimation atfogrdquo in Proceedings of the 2016 13th IEEE Annual ConsumerCommunications amp Networking Conference (CCNC) pp 12ndash17 IEEE Las Vegas NV USA January 2016

[56] N D Tung L L Bao and B Vijay ldquoPrice-based resourceallocation for edge computing a market equilibrium ap-proachrdquo IEEE Transactions on Cloud Computing p 1 2018

[57] X Xu C He Z Xu L Qi S Wan and M Z A BhuiyanldquoJoint optimization of offloading utility and privacy for edgecomputing enabled IoTrdquo IEEE Internet of Aings Journal2019

[58] X Xu Y Chen X Zhang Q Liu X Liu and L Qi ldquoAblockchain-based computation offloading method for edgecomputing in 5G networksrdquo Software Practice and Experi-ence 2019

[59] X Xu Y Li T Huang et al ldquoAn energy-aware computationoffloading method for smart edge computing in wirelessmetropolitan area networksrdquo Journal of Network and Com-puter Applications vol 133 pp 75ndash85 2019

[60] X Xu Y Xue L Qi et al ldquoAn edge computing-enabledcomputation offloading method with privacy preservation forinternet of connected vehiclesrdquo Future Generation ComputerSystems vol 96 pp 89ndash100 2019

14 Complexity

[61] G Yeting L Fang X Nong and C Zhengguo ldquoTask-basedresource allocation bid in edge computing micro datacenterrdquoComputers Materials amp Continua vol 61 no 2 pp 777ndash7922019

[62] X Chen L Jiao W Li and X Fu ldquoEfficient multi-usercomputation offloading for mobile-edge cloud computingrdquoIEEEACM Transactions on Networking vol 24 no 5pp 2795ndash2808 2016

[63] B Gao L He X Lu C Chang K Li and K Li ldquoDevelopingenergy-aware task allocation schemes in cloud-assisted mo-bile workflowsrdquo in Proceedings of the 2015 IEEE InternationalConference on Computer and Information Technology Ubiq-uitous Computing and Communications Dependable Auto-nomic and Secure Computing Pervasive Intelligence andComputing pp 1266ndash1273 IEEE Liverpool UK October2015

[64] X Xu X Zhang H Gao Y Xue L Qi andW Dou ldquoBecomeblockchain-enabled computation offloading for IoT in mobileedge computingrdquo IEEE Transactions on Industrial Infor-matics 2019

[65] W Yifei W Zhaoying G Da and Y F Richard ldquoDeepq-learning based computation offloading strategy for mobileedge computingrdquo Computers Materials amp Continua vol 59no 1 pp 89ndash104 2019

[66] M Barcelo A Correa J Llorca A M Tulino J L Vicarioand A Morell ldquoIoT-cloud service optimization in nextgeneration smart environmentsrdquo IEEE Journal on SelectedAreas in Communications vol 34 no 12 pp 4077ndash40902016

[67] B Cheng M Wang S Zhao Z Zhai D Zhu and J ChenldquoSituation-aware dynamic service coordination in an IoTenvironmentrdquo IEEEACM Transactions On Networkingvol 25 no 4 pp 2082ndash2095 2017

[68] V Angelakis I Avgouleas N Pappas and D Yuan ldquoFlexibleallocation of heterogeneous resources to services on an IoTdevicerdquo in Proceedings of the 2015 IEEE Conference onComputer Communications Workshops (INFOCOMWKSHPS) pp 99-100 IEEE Hong Kong China April 2015

[69] S Li Q Ni Y Sun G Min and S Al-Rubaye ldquoEnergy-ef-ficient resource allocation for industrial cyber-physical IoTsystems in 5g erardquo IEEE Transactions on Industrial Infor-matics vol 14 no 6 pp 2618ndash2628 2018

[70] X Liu Z Qin Y Gao and J A McCann ldquoResource allocationin wireless powered IoT networksrdquo IEEE Internet of AingsJournal vol 6 no 3 pp 4935ndash4945 2019

[71] W Ejaz and M Ibnkahla ldquoMulti-band spectrum sensing andresource allocation for IoT in cognitive 5g networksrdquo IEEEInternet of Aings Journal vol 5 no 1 pp 150ndash163 2017

[72] G Colistra V Pilloni and L Atzori ldquoTask allocation in groupof nodes in the IoT a consensus approachrdquo in Proceedings ofthe 2014 IEEE International Conference on Communications(ICC) pp 3848ndash3853 IEEE Sydney Australia June 2014

[73] J Li Q Sun and G Fan ldquoResource allocation for multiclassservice in IoTuplink communicationsrdquo inProceedings of the 20163rd International Conference on Systems and Informatics (ICSAI)pp 777ndash781 IEEE Shanghai China November 2016

[74] L I Zheng and K H Liu ldquoDynamic bandwidth resourceallocation algorithm in internet of things and its applicationrdquoComputer Engineering vol 38 no 17 pp 16ndash19 2012

[75] K Gai and M Qiu ldquoOptimal resource allocation using re-inforcement learning for IoT content-centric servicesrdquo Ap-plied Soft Computing vol 70 pp 12ndash21 2018

[76] X Xu W Dou X Zhang and J Chen ldquoEnreal an energy-aware resource allocation method for scientific workflow

executions in cloud environmentrdquo IEEE Transactions onCloud Computing vol 4 no 2 pp 166ndash179 2016

[77] K Bousselmi Z Brahmi and M M Gammoudi ldquoEnergyefficient partitioning and scheduling approach for scientificworkflows in the cloudrdquo in Proceedings of the 2016 IEEEInternational Conference on Services Computing (SCC)pp 146ndash154 IEEE San Francisco CA USA June 2016

[78] Y Sonia C Rachid K Hubert and G Bertrand ldquoMulti-objective approach for energy-aware workflow scheduling incloud computing environmentsrdquo Ae Scientific World Jour-nal vol 2013 Article ID 350934 13 pages 2013

[79] F Cao Efficient Scientific Workflow Scheduling in CloudEnvironment Southern Illinois University Carbondale ILUSA 2014

[80] Z Li J Ge H Hu W Song H Hu and B Luo ldquoCost andenergy aware scheduling algorithm for scientific workflowswith deadline constraint in cloudsrdquo IEEE Transactions onServices Computing vol 11 no 4 pp 713ndash726 2015

[81] M Khaleel and M M Zhu ldquoEnergy-aware job managementapproaches for workflow in cloudrdquo in Proceedings of the 2015IEEE International Conference on Cluster Computingpp 506-507 IEEE Chicago IL USA September 2015

[82] J Shi J Luo F Dong and J Zhang ldquoA budget and deadlineaware scientific workflow resource provisioning and sched-uling mechanism for cloudrdquo in Proceedings of the 2014 IEEE18th International Conference on Computer Supported Co-operative Work in Design (CSCWD) pp 672ndash677 IEEEHsinchu Taiwan January 2014

[83] Y Ge Y Zhang Q Qiu and Y-H Lu ldquoA game theoreticresource allocation for overall energy minimization in mobilecloud computing systemrdquo in Proceedings of the 2012ACMIEEE International Symposium on Low Power Electronicsand Design pp 279ndash284 ACM Redondo Beach CA USAAugust 2012

[84] X Wang Y Wang and C Yue ldquoAn energy-aware bi-leveloptimization model for multi-job scheduling problems undercloud computingrdquo Soft Computing vol 20 no 1 pp 303ndash3172016

[85] R Yanggratoke F Wuhib and R Stadler ldquoGossip-basedresource allocation for green computing in large cloudsrdquo inProceedings of the 2011 7th International Conference onNetwork and Service Management pp 1ndash9 IEEE ParisFrance October 2011

[86] A Paya and D C Marinescu ldquoEnergy-aware load balancingand application scaling for the cloud ecosystemrdquo IEEETransactions on Cloud Computing vol 5 no 1 pp 15ndash272015

[87] V D Justafort R Beaubrun and S Pierre ldquoA hybrid ap-proach for optimizing carbon footprint in intercloud envi-ronmentrdquo IEEE Transactions on Services Computing vol 12no 2 pp 186ndash198 2016

[88] R Panwar and B Mallick ldquoLoad balancing in cloud com-puting using dynamic load management algorithmrdquo inProceedings of the 2015 International Conference on GreenComputing and Internet of Aings (ICGCIoT) pp 773ndash778IEEE Noida India October 2015

[89] C-T Yang K-C Wang H-Y Cheng C-T Kuo andW C C Chu ldquoGreen power management with dynamicresource allocation for cloud virtual machinesrdquo in Proceedingsof the 2011 IEEE International Conference on High Perfor-mance Computing and Communications pp 726ndash733 IEEEAlberta Canada September 2011

[90] C-T Yang H-Y Cheng and K-L Huang ldquoA dynamicresource allocation model for virtual machine management

Complexity 15

on cloudrdquo in Proceedings of the International Conference onGrid and Distributed Computing pp 581ndash590 SpringerGangneug Korea December 2011

[91] X Xu F Shucun C Qing et al ldquoDynamic resource allocationfor load balancing in fog environmentrdquo Wireless Commu-nications amp Mobile Computing vol 2018 no 2 Article ID6421607 15 pages 2018

[92] J Oueis E C Strinati and S Barbarossa ldquoe fog balancingload distribution for small cell cloud computingrdquo in Pro-ceedings of the 2015 IEEE 81st Vehicular Technology Confer-ence (VTC Spring) pp 1ndash6 IEEE Glasgow UK May 2015

[93] K Wang Y Wang Y Sun S Guo and J Wu ldquoGreen in-dustrial internet of things architecture an energy-efficientperspectiverdquo IEEE Communications Magazine vol 54 no 12pp 48ndash54 2016

16 Complexity

the least amount of power He et al [8] proposed a service-based system supporting keyword search in which dif-ferent search keywords represent different tasks ismethod can help unprofessional service users build ser-vice-oriented systems Sun et al [9] proposed a cloudservice selection method to measure and aggregate thenonlinear relationship between standards And a frame-work based on priority is designed to determine the criteriarelationships and weights when historical information isinsufficient Mazzucco and Dyachuk [10] were also com-mitted to making cloud service providers obtain the largestprofits ey proposed the dynamic distribution strategy ofpowering server switch e strategy not only enables usersto get good service but also reduces power consumptione number of live servers determines the state of thesystem but running or closing a server cannot be done in aflash So it is important to take into account of the timeGiven the short time required for the server switch for-mula (1) [10] represents the cost of changing the number ofservers running per unit time In order to make users havea good experience this paper further uses a forecastingmethod to accurately predict the usersrsquo time-changingneeds

Q Δnt

1113944

l

i1di + kre3

⎛⎝ ⎞⎠ (1)

where t represents the observation time Δn represents thenumber of servers whose state change over time di

represents the cost of the state change of a hardwarecomponent e3 represents the energy consumed in a unittime to change the state k represents the average time tochange the state of a server and l represents the amount ofcomponent

Mazzucco et al [7] and Mazzucco and Dyachuk [10]both explore strategies of reducing the power cost of runninga data center and changing the on-off state of the serversetwo strategies can maximize the usersrsquo experience and saveenergy at the same time eir difference is that Mazzuccoand Dyachuk [10] believes that it is impossible to accuratelypredict the changes of usersrsquo needs over time So comparedwith paper [7] the strategy proposed in paper [10] is fault-tolerant He et al [11] proposed three service selectionmethods that support QoS and can combine multitenantservice-based systems ese three methods can achievethree degrees of multitenant maturity which is more effi-cient than the traditional single-user approach Sun et al[12] proposed a unified semantic model to describe cloudservice is model expands the basic structure of unifiedservice description language And it defines a transactionmodule to model the rating system for cloud services fromvarious perspectives So it can improve the ability of themodel on service ranking In addition an annotation systemis put forward to enrich the language expression Wang et al[13] proposed a fault-tolerant strategy based on multitenantservice criticality which can provide redundancy for keycomponent services evaluating the criticality of eachcomponent service to determine the optimal fault-tolerantpolicy erefore the quality of the multitenant based ser-vice system can be guaranteed Mustafa et al [14] leveragedthe notion of workload consolidation to improve energyefficiency by putting incoming jobs on as few servers aspossible e concept of SLA is also imported to minimizethe total SLA violations Given that a change in workloadchanges the utilization of CPU required over time So anintegral function (formula (2) [14]) is used to represent thetotal energy (E) consumed by a server (S) operation

E 1113946t1

t0P(u(t))dt (2)

where P is the amount of power consumed by the server interms of CPU utilization (u) in time t

Bi et al [15] established an architecture that can ad-ministrate itself in cloud data centers firstlye architectureis suitable for web application services with several levels andhas virtualization mechanismen a mixed queuing modelis proposed to decide the number of virtual machines (VMs)in each layer of application service environments Formula(3) [15] is used to represent the local profit that can be madeby the ith virtualization application service environmentFinally a problem of misalignment restrained optimizationand a heuristic mixed optimization algorithm are proposedBoth of them can make more revenues and meet require-ments of different customers

Pi(E) Revenue(E) + Penalty(E) + Loss(E) + Cost(E)

(3)

where Revenue(E) Penalty(E) Loss(E) and Cost(E) re-spectively represent the total benefit penalty loss and costof VMs

Singh et al [16] proposed a technology named STARwhich can manage resources itself in the cloud computing

Data

Data

Data producer

Data producerconsumer

RequestResult

Edge

Computing offloadData cachingstorage

data processingRequest distribution

Service deliveryIoT management

Privacy protection

Figure 3 Structure of edge computing [2]

Complexity 3

environment and reduce SLA violations So the paymentefficiency of cloud services can be improved Beloglazov andBuyya [17] proposed a system to manage energy in the clouddata center By continuously integrating VMs and dy-namically redistributing VMs the system can achieve thegoal of saving energy and providing a high QoS level at thesame time Guazzone et al [18] proposed an automaticmanagement system (see Figure 4) for resources to providecertain QoS levels and reduce energy consumption Re-source manager of the framework in Figure 4 combinesvirtualization technologies and control-theoretic technolo-gies Virtualization technologies deploy each application toindependent VM And control-theoretic technologies realizethe automatic management of computer performance andenergy consumption In addition the resource managerconsists of several independent components named Ap-plication Manager Physical Machine Manager and Mi-gration Manager Different from traditional static methodsthis method can both fit the changing workloads dynami-cally and achieve remarkable results in reducing QoS vio-lations Sun et al [19] established a model to simplify thedecision of cloud resource allocation and realize the inde-pendent allocation of resources e optimal resourceconfiguration can be obtained so the QoS requirements canbe well met Siddesh and Srinivasa [20] explored theproblems of dynamic resource allocation and SLA assuranceey proposed a framework to deal with heterogeneousworkload types by dynamically planning computing capacityand assessing riskse framework uses scheduling methodsto reduce SLA violation risks and maximize revenues inresource allocation

Garg et al [21] proposed a resource allocation strategyfor VM dynamic allocation e strategy can improve re-source utilization increase providersrsquo profits and reduceSLA violations Jing et al [22] proposed a new dynamicallocating technique using the mixed queue model meetingcustomersrsquo different requirements of performance by pro-viding virtualized resources to each layer of virtualizedapplication services All these methods can reasonablyconfigure resources in the cloud data center improve systemperformance reduce additional costs of using resources andmeet the required QoS

Qi et al [23] proposed a QoS-aware VM schedulingstrategy named QVMS to satisfy QoS Firstly the schedulingproblem is transformed into a problem with several ob-jectives And then the optimal VM migration method isfound according to the genetic algorithm e schedulingstrategy can effectively manage resources in the networkphysical system thus reducing the energy consumption andimproving QoS levels Qi et al [25] proposed a servicerecommendation strategy by considering the time factor toimprove the traditional location-sensitive hash technologye policy emphasizes the influence of dynamic factors onQoS and the protection of user privacy

Table 1 shows a summary of the abovementioned workse solution of the problems in Table 1 can improve QoS incloud computing environment Server management refers todynamically allocating powering serversrsquo switches Work-loads consolidation refers to combining work to save energy

VM management refers to reasonable scheduling or inte-gration of VMs to achieve better performance Self-man-agement refers to the realization of self-management ofresources which can achieve higher efficiency Resourcemanagement refers to the correct allocation of resources toreduce waste Service management is about making rea-sonable service choices [26]

212 Fog Computing Gu et al [27] used fog computing toprocess a large amount of data generated by medical devicesand built Fog Computing Supported Medical Cyber-Phys-ical System (FC-MCPS) In order to reduce the cost of FC-MCPS research studies were carried out on the joint of basestation task assignment and VM layout e problem ismodeled as a mixed integer linear programming (MILP) Atwo-stage heuristic algorithm based on linear programming(LP) is proposed to solve the problem Ni et al [28] proposeda resource allocation approach based on fog computingwhich enables users to select resources independently Inaddition this approach takes into account the price and timerequired to finish the job Formula (4) [28] is used to definethe credibility BCreij of Resource Rj received from useriwhen the user interacts with Resource Ri

BCreij ω1λresp + ω2cexec + ω3 1 minus ηreboot( 1113857 + ω4μrel

(4)

where the value of ωkϵ[0 1] 11139364k1 ωk 1 which can be

determined by the user or the actual situation λresp cexecηreboot and μrel are the response speed of the corre-sponding index service the efficiency of execution the speedof restart and the reliability respectively

213 Edge Computing Wei et al [29] proposed a unifiedframework in the sustainable edge computing to save energyincluding the energy that is distributed and renewable Andthe architecture can combine the system that supply energyand edge services which can make full use of renewableenergy and provide better QoS Lai et al [30] proposed anoptimized allocation method for edge usersemethod cannot only maximize the amount of resources allocated tousers but also consider the dynamic QoS level of users Soedge user allocation problem can be made more general andimproving the quality of experience

214 MEC Xu et al [31] used block chain to improve thetraditional crowdsourcing technology Firstly they proposeda mobile crowdsourcing framework using block chaintechnology to protect user privacy en they used dynamicprogramming strategy of clustering algorithm to classifyrequesters Finally they generated service policies to balanceprofits and energy consumption

215 IoT Rolik et al [32] proposed a method to build aframework of IoT infrastructure based on microcloud emethod can help users use resources rationally reduce thecost of managing infrastructure and improve the quality of

4 Complexity

life of consumers He et al [33] proposed a dynamic networkslice strategy e network slice can be dynamically adjustedaccording to the time-varying resource demands is methodcan improve the utilization of the underlying resources andbetter meet different QoS demands Yao and Ansari [34]proposed an algorithm to determine the number of VMs to berented and to control the power supply us the cost of thesystem can be minimized and the QoS can be improvedFormula (5) [34] is used to limit the delay requirement of QoSe total delay must not exceed the computation deadline ofeach task and the total delay is composed of wireless trans-mission delay and fog processing delay

tci + t

wi leDi foralliϵN (5)

where c and w respectively represents fog processing andwireless transmission i denotes a location tc

i represents thedelay of processing tw

i represents the delay of wirelesstransmission Di denotes the deadline and N denotes dif-ferent locations

22 Resource Management and Allocation Rational alloca-tion of resources is an effective means to save energy

221 Cloud Computing Wang et al [35] introduced anallocation method based on distributed multiagent to al-locate VMs to physical machines e method can realize

VM consolidation and consider the migration costs si-multaneously In addition a VM migration mechanismbased on local negotiation is proposed to avoid unnecessaryVM migration costs Hassan et al [37] established a for-mulation of universal problem and proposed a heuristicalgorithm which has optimal parameters Under this for-mulation dynamic resource allocation can be made to meetthe QoS requirements of applications And the cost neededfor dynamic resource allocation can be minimized with thisalgorithm Wu et al [38] proposed a scheduling algorithmbased on the technology that can scale the voltage frequencydynamically in cloud computing e algorithm can allocateresources for performing tasks and realize low powerconsumption network infrastructure Compared with otherschemes this scheme not only sacrifices the performance ofexecution operations but also saves more energy

Sarbazi and Zomaya [45] used two job consolidationheuristic methods to save energy One is MaxUtil to betterutilize resources and the other is Energy-Conscious TaskConsolidation to reduce energy consumption ese twomethods can promote the concurrent execution of multipletasks and improve the energy efficiency Hsu et al [46]proposed a job consolidation technique to minimize energyconsumption Formula (6) [46] defines the energy con-sumption of VM Vi from time t0 to tm in the cluster isdefined And formula (7) [46] defines the total energyconsumption in a virtual cluster VCk in the period Inaddition the proposed technique limits the CPU usage and

Reference machine

Reference machine

SLA

SLA

Application A1

Application An

TIER 1

TIER 2

TIER 3

TIER 1

TIER 2

TIER 3

Cloud infrastructure

VM

VM

VM

VM

VM

VM

Performance measuresEnergy consumptions

Resource costs

Resourcemanager

Migrationmanager

Physicalmachinemanagers

Applicationmanagers

Figure 4 Framework of the proposed three-fold system [18]

Complexity 5

Tabl

e1

Worksummaryof

QoS

guaranteeing

orSL

Aassurancein

clou

dcompu

ting

Subp

roblem

sSolutio

nsLiteratures

Advantages

Server

managem

ent

Apo

licyof

dynamicallocatio

nof

poweringserverss

witches

[71

0]Maxim

izes

benefitsim

proves

QoSa

ndminim

izes

power

consum

ption

Workloads

consolidation

Atechniqu

eforconsolidatingworkloads

[14]

Achievesenergy

saving

sby

usingfewestserverswhile

redu

cing

SLA

violations

VM

managem

ent

Anarchitecturethat

canadministrate

itselfa

ndamixed

queuingmod

el[15]

Makes

morerevenu

esandmeets

different

requ

irem

ents

ofcustom

ersanddecidesthe

numberof

VMsforeach

layerof

avirtuala

pplication

Asystem

tointegrateanddynamically

redistribu

teVMs

[17]

Achievesthe

goalofsaving

energy

throughVM

integrationandprovidesah

ighQoS

levelat

thesametim

e

AQoS-awareVM

schedu

lingstrategy

named

QVMS

[23]

Effectiv

elymanages

resourcesin

thenetworkph

ysical

system

toredu

cetheenergy

consum

ptionandim

proves

QoS

AVM

integrationmetho

dwith

severaltargets

[24]

Savesenergy

andredu

cesSL

Aviolations

byapplying

different

strategies

todifferent

load

states

oftheho

st

Self-managem

ent

Atechno

logy

named

STAR

[16]

Redu

cesSL

Aviolations

andim

proves

paym

ente

fficiency

ofclou

dservices

Adynamic

resource

managem

entsystem

[18]

Self-manages

theresourceso

fcloud

infrastructurestoprovidea

ppropriateQoS

andfitsthe

changing

workloads

dynamically

Resource

managem

ent

Amod

elthat

canrealizetheindepend

entallocatio

nof

resources

[19]

Obtains

theop

timal

resource

confi

guratio

nmeets

theQoS

requ

irem

entsand

provides

econ

omical

clou

dresources

Adynamic

resource

allocatio

nstrategy

[20

21]

Redu

cesSL

Aandmaxim

izes

revenu

esandresource

utilizatio

non

theclou

d

Amixed

queuemod

el[22]

Reason

ably

confi

gurestheresourcesin

theclou

ddata

centerimproves

thesystem

performancereduces

theadditio

nalcosto

fusin

gresourcesmeetstherequ

ired

QoSand

provides

virtualresou

rces

toeach

layerof

virtuala

pplicationservices

Service

managem

ent

Aun

ified

semantic

mod

elthat

candescribe

clou

dservice

[12]

Improves

theability

ofmod

elon

servicerank

ingandenriches

thelang

uage

expressio

n

Arecommendatio

nservicestrategy

[25]

Emph

asizes

theinflu

ence

oftim

efactorson

QoS

andim

proves

thetradition

allocatio

n-sensitive

hash

techno

logy

toprotectusersrsquoprivacy

Aclou

dserviceselectionmetho

dusingfuzzymeasure

and

Cho

quet

integral

andafram

eworkbasedon

priority

[9]

Selectsservicewhenhistorical

inform

ationisinsufficientto

determ

inethecriteria

relatio

nships

andweigh

ts

reeserviceselectionmetho

dsthat

supp

ortQ

oSandcan

combine

multitenants

ervice-based

system

s[11]

Achievesthree

degreeso

fmultitenantm

aturity

which

ismoreeffi

cientthanthetradition

alsin

gle-user

approach

Afault-tolerant

strategy

basedon

multitenants

ervice

criticality

[13]

Guaranteesthequ

ality

ofthemultitenant-basedservicesystem

Aservice-basedsystem

supp

ortin

gkeyw

ordsearch

[8]

Effectiv

elyhelpssystem

engineerswho

areno

tfam

iliar

with

service-oriented

architecture

techno

logy

tobu

ildservice-oriented

system

s

6 Complexity

merges tasks in virtual clusters Once a task migrationhappens the energy cost model will take into account thenetwork latency Sarbazi-Azad and Zomaya [45] and Hsuet al [46] both maximize the benefit of cloud resources byusing task merging techniques Sarbazi-Azad and Zomaya[45] uses a greedy algorithm called MaxUtil While Hsuet al [46] takes into account the network latency associatedwith task migration So in [46] a 17 improvement isachieved over MaxUtil

E0m Vi( 1113857 1113944m

t0Et Vi( 1113857 (6)

E0m VCk( 1113857 1113944n

i0E0m Vi( 1113857 (7)

where Et is the energy consumption in unit time and n is thenumber of VMs in the cluster

Hsu et al [47] proposed a task integration technologybased on the energy perception According to the charac-teristics of most cloud systems the principle of using 70CPU is proposed to administrate job integration amongvirtual clusters is technology is very effective in reducingthe amount of energy consumed in cloud systems bymerging tasks Panda and Jana [48] proposed an algorithmwith several criteria to combine tasks e algorithm notonly considers the time needed for processing jobs but alsoconsiders the utilization rate of VMs And the algorithm ismore energy efficient because it takes into account not onlythe processing time but also the utilization rate of VMsWang and Su [39] proposed a resource allocation algorithmto deal with wide range of communication between nodes incloud environment is algorithm uses recognition tech-nology to dynamically distribute jobs and nodes accordingto computing ability and factors of storage And it can re-duce the traffic when allocating resources because it usesdynamic hierarchy Lin et al [40] proposed a dynamicauction approach for resource allocation e approach canensure that even if there are many users and resourcesproviders will have reasonable profits and computing re-sources will be allocated correctly Yazir et al [41] proposeda new method to manage resources dynamically and au-tonomously Firstly resource management is split into jobsand each job is executed by autonomous nodes Secondautonomous nodes use the method called PROMETHEE toconfigure resources Krishnajyothi [36] proposed a frame-work which can implement parallel task processing to solvethe problem of low efficiency when submitting large tasksCompared with the static framework this framework candynamically allocate VMs thus reducing costs and the timeof processing tasks Lin et al [42] proposed a method toallocate resources dynamically by using thresholds Becausethis method uses the threshold value it can optimize thereallocation of resources improve the usage of resourcesand reduce the cost Xu et al [43] proposed a data placementstrategy named IDP for the data generated by IoTs devices toachieve reasonable data placement In this way the privacyof these data can be protected while resources are allocatedreasonably Jo et al [44] proposed a computing offload

framework under 5G network e framework transfers thecomputing burden to the cloud thus reducing the com-puting load of clients and the communication cost

Table 2 shows a summary of the abovementioned workse problem of resource allocation and management incloud computing can be divided into problems in Table 2VM management is about a reasonable configuration ofVMs Resource allocation represents the dynamic andflexible allocation of resources Task integration refers tocombining tasks to save energy and improve efficiency

222 Fog Computing Yin et al [49] established a newmodel of scheduling jobs which applies containers In orderto make sure that jobs can be finished on time a jobscheduling algorithm is developed e algorithm can alsooptimize the number of tasks that can be performed togetheron the nodes in fog computing And this paper proposes aredistribution mechanism to shorten the delay of tasksese methods are very effective in reducing task delaysAazam and Huh [50] established a framework to admin-istrate resources effectively in the mode of fog computingConsidering that there are various types of objects anddevices the connection between them may be volatile So amethod for predicting and administrating resources isproposed e method considers that any object or devicecan quit using resources at anytime Cuong et al [5] studiedthe allocating resources jointly problem and the problem ofcarbon footprint minimization in fog data center Formula(8) [5] is used to denote the energy consumption of serversIn addition a distributed algorithm is proposed to solve theproblem of wide range optimization

P(y) C middot Pidle + Ppeak minus Pidle1113872 1113873 middot y middot κ (8)

where P(y) represents the power supply required by theservers in a data center y represents the video stream κdenotes a conversion factor that converts the video streaminto workload C represents the data centerrsquos load capacityand Pidle and Ppeak respectively represent the idle power andpeak power of the servers

Jia et al [51] studied the problem of computing resourceallocation in fog computing network with three levelsFirstly the problem of resource allocation is transformedinto a bilateral matching optimal problem And then a bi-matching approach is proposed for this problem which canimprove the performance of the system and obtain highercost efficiency Zhang et al [52] proposed a framework forjoint optimization under fog computing to allocate fognodesrsquo finite computing resources e framework canachieve the best allocation and effectively improve thenetworksrsquo performance Tan et al [53] presented a method toallocate computing and communication resources emethod transfers computing jobs to remote cloud and nodesand simplifies edge nodesrsquo computing and computing en-ergy Vasconcelos et al [54] developed a platform to allocateresources accessible to client devices in fog computingenvironment allocating the resources of devices near thehost to meet the applications needs for rapid response tocomputing resources Aazam et al [55] presented a method

Complexity 7

to estimate and manage resource in fog computing emethod is based on the fluctuation of customer abandon-ment probability type and price of service and so on

Table 3 shows a summary of the abovementioned workse problems in Table 3 are also derived from resourceallocation and management problems Task allocationrepresents the scheduling and redistribution of tasks Re-source allocation is still about the dynamic and flexibleallocation of resources Low latency refers to taking shorttime to configure and manage resources which can improveefficiency

223 Edge Computing Tung et al [56] proposed a newframework for resource allocation based on market needse resources come from edge nodes (ENs) with limitedheterogeneous capabilities and are allocated to multiplecompeting services on the network edge Generating amarket equilibrium solution by reasonably pricing ENs canobtain the maximum utilization of marginal computingresources Xu et al [57] proposed a strategy to optimizeoffloading and privacy protection is strategy shifts tasksfirstly to improve the resource utilization of resource-limited

edge cells And then it balances QoS performance andprivacy protection to achieve joint optimization

Xu et al [58] proposed an offload strategy for edgecomputing under 5G network which uses block chaintechnologye optimal strategy is further obtained by usingthe balanced offloading method It solves the problem ofdata loss under the condition of transmission delay which iscaused by the uneven requirements of user equipments onresources Xu et al [59] proposed a computational off-loading method named EACO to reduce the energy con-sumption in smart computing models Figure 5 showsarchitecture of smart edge computing where the shortestpath is used to unload tasks EACO uses genetic algorithmsto reduce the energy consumption for operating edgecomputing nodes and improve the efficiency of performingcomplex computing tasks Xu et al [60] proposed a com-putational offloading strategy for edge computing to protectthe privacy of interconnected vehicle networks ey firstlyanalyzed privacy conflicts of tasks And then they designedthe communication route to obtain routing vehicles whichcan achieve the optimization of several objectives Yetinget al [61] proposed a unique resource allocation mechanisme mechanism takes each individual task as the basis for

Table 2 Work summary of resource allocation and management in cloud computing

Subproblems Solutions Literatures Advantages

VMmanagement

A VM allocation method based on severaldistributed agents [35] Centralizes VMs to physical machines and reduces the

overall energy costA framework which can implement parallel

task processing [36] Dynamically allocates VMs for large tasks

Resourceallocation

A general problem formula and a heuristicalgorithm with optimization parameters [37]

Dynamically allocates resources according to QoSrequirements and realizes energy saving by optimizing the

number of serversA scheduling algorithm based on thetechnology that can scale the voltage

frequency dynamically[38] Ensures the performance of executing jobs while

implementing green computing

A fuzzy pattern recognition technology [39] Reduces the traffic when allocating resources and usesdynamic hierarchy

A dynamic auction approach [40]Guarantees profits of providers and allocates resourcescorrectly when there are a large amount of users and

resources

PROMETHEE [41]Allows node agents to decompose and execute tasksautonomously to improve the flexibility of resource

allocation

A method using thresholds [42]Optimizes resource reallocation improves resource usagereduces cost and studies resource allocation strategies at the

application level

IoT-oriented data placement (IDP) method [43] Protects data privacy allocates resources reasonably andfocuses on the placement method of IoT data

A computing offload framework under 5Gnetwork [44] Reduces the computing load of clients and the

communication cost

Taskconsolidation

Two job consolidation heuristics namedldquoMaxUtilrdquo and energy-conscious task

consolidation[45] Promotes concurrent execution of multiple tasks and

improves energy efficiency

A job consolidation technique aiming atenergy saving [46 47]

Reduces the power consumption of cloud system protectsdata privacy allocates resources reasonably limits CPU

usage and merges tasks in virtual clusters

An algorithm with several criteria [48]Takes into account the job processing time and the VM

utilization rate simultaneously Dramatically reduces energyconsumption compared with state-of-the-art works

8 Complexity

resource allocation rather than for the whole service It reducesthe packet loss rate and saves energy by unloading services

224 MEC Chen et al [62] studied the problem of com-puting unloading with several users in the environment ofMEC with wireless interference which have many channels Inaddition a distributed algorithm for computing unloading isdevelopede algorithm can perform the unloading well evenwhen there are a large number of users Gao et al [63] built aquadratic binary program which is able to assign tasks inmobile cloud computing environment Two algorithms arepresented to obtain the optimal solution Both of these heuristicalgorithms can effectively solve the task assignment problemXu et al [64] proposed an offloadingmethod using block chaintechnology It can guarantee the loss of data in offloading tasksunder edge computing And it can solve the problem of re-source requests out of proportion due to the limited load of

edge computing equipment during task transfer Yifei et al [65]proposed a model-free reinforcement learning framework tosolve the problem of computational unloadingis model canbe applied to the computational unloading with time-changingcomputing requests

225 IoT Barcelo et al [66] expressed the problem ofservice allocation [67] as a mixed flow problem with min-imum cost which can be solved by LP solving this serviceallocation problem can solve the problems of unbalancednetwork load and delay of end-to-end service And it canalso figure out the problem of excessive consumption ofelectricity brought by the architecture of centralized cloudAngelakis et al [68] assigned the requirements of servicesresources to heterogeneous network interfaces of equipmentsSo more heterogeneous network interfaces can be used by alarge amount of services

Mobile devices

Mobile devices ECN

ECN

ECNECN

AP

AP

AP

APAP

Figure 5 An architecture of smart edge computing [59]

Table 3 Work summary of resource allocation and management in fog computing

Problems Solutions Literatures AdvantagesTaskallocation

A new model of scheduling jobs and aredistribution mechanism [49] Finishes jobs on time optimizes the number of tasks and

shortens the delay of tasks

Resourceallocation

A framework to administrate resources and amethod to predict and administrate resources [50]

Assists service providers to predict the amount of availableresources based on different types of service customers anddeals with the phenomenon that objects or devices withdraw

from resource utilization at any time

A bi-matching approach [51] Improves the performance of the system and obtains highercost efficiency

A framework for joint optimization [52] Optimizes resource allocation and improves networkperformance

A method to allocate computing andcommunication resources [53]

Allocates computing and communication resources transferscomputing jobs to remote cloud and nodes simplifies edge

nodesrsquo computing and saves computing energyA platform to allocate resources accessible to

client devices [54] Enables rapid response to computing resources and allocatesthe resources of devices near the host

A framework to manage resources [55] Considers the fluctuation of the customer abandonmentprobability and service types

Low latency A distributed algorithm [5] Solves the problem of wide range optimization and allocatesresources jointly

Complexity 9

Li et al [69] proposed communication framework in 5Gand studied the problem of allocating power andchannels So the signal data in the channel can beavailable and the total energy efficiency can be maximumFormula (9) [69] shows how to calculate the energy ef-ficiency of a system

U 1113944K

k11113944

M

i1EES

iK + 1113944K

K11113944

N

i1EEA

jk (9)

where EESiK and EEA

jk respectively denote the energy ef-ficiency of sensor S and actuator A on channels e sets ofsensors actuators and channels are respectively repre-sented as s S1 S2 SM1113864 1113865 A A1 A2 AN1113864 1113865 andC C1 C2 CK1113864 1113865

Liu et al [70] studied the problem of allocating resourcesefficiently on IoT that supplies wireless power In thismethod users are first grouped into accessible channelsAnd then power distribution of users grouped in the samechannel is studied to improve throughput of the networkis method can allocate finite resources to a large group ofusers Ejaz and Ibnkahla [71] proposed the resource allo-cation framework with several bands under cognitive 5GIoT In the highly dynamic environment of the IoTmultiband method can manage resources more flexiblyand reduce more energy consumption In addition areconstruction approach with several levels is proposed toallocate resources reasonably for applications with dif-ferent needs of QoS Colistra et al [72] proposed aprotocol which is distributed and optimal to allocateresources in heterogeneous IoT Because this protocol hasexcellent adaptability when changing topology of net-work it can distribute resources evenly among nodes Jianet al [73] proposed a multilevel allocating resources al-gorithm for IoT communication using advanced tech-nology e algorithm uses hierarchical structure and hasfast data processing rate and very low latency in a satu-rated or not saturated environment

Zheng and Liu [74] proposed a new algorithm to allocatebandwidth dynamically for controlling remote computers inthe IoT is method can reduce the error of signal re-construction under the same bandwidth and make thebandwidth allocation of IoT more reasonable Gai and Qiu[75] used reinforcement learning mechanisms to allocateresources to achieve high Quality of Experience ismethod can effectively solve the resource allocation prob-lems caused by the mismatch of service quality and complexservice providing condition in the IoT

Table 4 shows a summary of the above works eproblem in Table 4 represents the realization of dynamic andflexible allocation of resources e resources here canrepresent channels bandwidth and power

23 Scientific Workflow Execution Implementing scientificworkflows especially in heterogeneous environments canreduce resource waste and reduce energy costs Scientificworkflow can be obtained by reasonably allocating resourcesand dynamically deploying VMs

231 Cloud Computing Xu et al [76] proposed a resourceallocation method called EnRealan to solve the problem ofenergy consumption e dynamic deployment of VMs isgenerally adopted to execute scientific workflows Bousselmiet al [77] proposed a scheduling method based on energyperception for executing scientific workflows in cloudcomputing At first an algorithm of splitting workflows forenergy minimization is presented which can achieve a highparallelism without huge energy consumption en aheuristic algorithm used to optimize cat swarm is proposedfor the created partitions e algorithm can minimize thetotal consumption of energy and the execution time ofworkflows Sonia et al [78] proposed a workflow schedulingmethod with several objects and hybrid particle swarmoptimization algorithm In addition a method for dy-namically scaling voltage and frequency is proposed emethod canmake the processors work at any voltage level soas to minimize the energy consumption in the process ofworkflow scheduling Both Bousselmi et al [77] and Soniaet al [78] use scheduling method to achieve scientificworkflows and study the problem of energy consumptione difference is that Bousselmi et al [77] focuses on in-tensive computing tasks while Sonia et al [78] focuses onworkflow scheduling on heterogeneous computing systems

Cao [79] established a scheduling algorithm of scientificworkflows with an objective of energy savingis algorithmcan enable service providers to gain high profits and reduceusersrsquo overhead at the same time Li et al [80] proposed ascheduling algorithm based on cloud computing which canminimize cost of performing workflows within a specifiedtime In addition the rented VM was modified to save costfurther Khaleel and Zhu [81] proposed a scheduling algo-rithm and took scientific workflows as a model to make fulluse of cloud resources and save energy Shi et al [82]designed a flexible resource allocation and job schedulingmechanism to implement scientific workflows Because thismechanism can implement scientific workflows withinprescribed budgets and deadlines so it can work better thanother mechanisms

Table 5 shows a summary of the abovementioned workse problems in Table 5 are derived from the imple-mentation of the scientific workflow VM deployment refersto the rational allocation of VMs Workflow scheduling refersto reducing the scheduling energy and time In addition itrefers to the scheduling of the workflow on heterogeneoussystems Cost reduction refers to reducing the cost of workflowexecution Effective implementation is about scientific work-flow execution within a specified budget and time

24 Server Optimization Server optimization is also a goodway to save energy e goal of optimizing the server can beachieved by uninstalling unnecessary servers or consoli-dating servers as well as by reasonably scheduling tasksUnlike QoS optimization server optimization aims to op-timize the number of used servers improves the energyefficiency of servers and consolidates servers However QoSoptimization studies how tomake users get better experienceand meet their needs

10 Complexity

241 Cloud Computing Ge et al [83] proposed a game-theoretic method and transformed the problem of mini-mizing energy into a congestion game All mobile devices inthis method are participants in the game e methodchooses a server to unload the computation tasks to optimizeQoS levels and save energy which can optimize the systemand save energy Wang et al [84] proposed a MapReduce-based multitask scheduling algorithm to achieve the ob-jective of energy saving is model is a two-layer modelwhich considers the impact of server performance changeson energy consumption and the limitation of networkbandwidth In addition a local search operator is designedbased on which a two-layer genetic algorithm is proposede algorithm can schedule tens of thousands of tasks incloud and achieve large-scale optimization Yanggratoke

et al [85] proposed a general generic gossip protocol aimingat allocating resources in cloud environment An instanti-ation of this protocol was developed to enable server con-solidation to allocate resources to more servers to meetchanging load patterns

25 LoadBalancing Load balancing can help save energy bymanaging the number of servers and allocating resources

251 Cloud Computing Paya and Marinescu [86] intro-duced an operation model that balances cloud computingload and expands applications to save energy e principleof this model is to define an operating system e systemshould make servers run in the system as many as possible

Table 5 Work summary of scientific workflow execution in cloud computing

Problems Solutions Literatures Advantages

VM deployment A resource allocation method namedEnRealan [76] Performs scientific workflows based on energy perception

across cloud platforms

Workflowscheduling

A scheduling method based on energyperception [77]

Achieves a high parallelism without huge energyconsumption and minimizes the total consumption of

energy and execution time of workflows

A workflow scheduling method with severalobjects and hybrid particle swarm

optimization algorithm[78]

Makes the processors work at any voltage level minimizesthe energy consumption in the process of workflowscheduling and studies the scheduling problem of

workflows on heterogeneous systemsA scheduling algorithm based on various

applications [79] Enables service providers to gain high profits and reducesuser overhead at the same time

Cost reduction A scheduling algorithm based on energyperception [80 81] Minimizes the cost of performing workflows while

meeting the time constraintEffectiveimplementation

A flexible resource allocation and jobscheduling mechanism [82] Implements scientific workflows within prescribed

budgets and deadlines

Table 4 Work summary of resource allocation and management in IoT

Problems Solutions Literatures Advantages

Resourcemanagement

A distributed cloud network framework [66]Replaces the centralized architecture with a distributed cloud

architecture solves the defects of the centralized cloudarchitecture and brings people better experience

A MILP model [68] Assigns the requirement of servicesrsquo resource to heterogeneousnetwork equipment interface

A framework for communication used in5G [69]

Transforms the resource allocation problem into a power andchannel allocation problem minimizes the total energy

consumption and improves QoS levelsA low complexity channel allocation

algorithm [70] Improves throughput of the network and allocates finiteresources to a large group of users

A resource allocation framework withseveral bands under cognitive 5G IoT [71] Manages resources more flexibly and reduces energy

consumption than common single-band approach

A protocol which is distributed andoptimal to allocate resources [72]

Has excellent adaptability when changing topology of networkand dynamically manages resources in the heterogeneous IoT

environmentA multilevel allocating resourcesalgorithm for IoT communication [73] Has fast data processing rate and very low latency in both

saturated and nonsaturated environment

A new algorithm to allocate bandwidthdynamically [74]

Reduces the error of signal reconstruction under the samebandwidth and makes the bandwidth allocation of IoT more

reasonable

A reinforcement learning mechanism [75]Effectively solves the resource allocation problems caused by themismatch of service quality and complex service providing

condition in the IoT

Complexity 11

When no tasks are being performed the system should adjustservers to sleep thus energy consumption can be reducedJustafort et al [87] mainly studied the problem of workloaddistribution across cloud computing environment and pro-posed a method to solve the problem of the VM layout So thefootprint of carbon can be effectively reduced Panwar andMallick [88] proposed an algorithm to dynamically manage theload and effectively distribute the total incoming requestsbetween VMs rough efficient and uniform utilization ofresources this algorithm can achieve uniform distribution ofload between servers Yang et al [89] proposed a powermanagement mechanism to balance the load e system canmonitor VMs and dynamically allocate the resources Yanget al [90] proposed an optimization system to better allocateresources dynamically which can balance the load of VMsrunning on multiple physical machines Under this systemVMs can be migrated automatically to adjust high and lowloads without interrupting services Yang et al [89 90] manageVMs to achieve load balancing ey allocate resources dy-namically to migrate VMs which can balance workloads ondifferent physical machines e difference is that Yanget al[89] integrate a dynamic resource allocation approachwithOpenNebulaWhile Yang et al [90] focuse on avoiding serviceoutages during VM migration

Table 6 shows a summary of the abovementioned workse problems in Table 6 are from the load balancing problemServer management is about the control of the number ofservers running in the system Workload management is therational allocation of workload or tasks VMmanagement refersto configuringVMresources andmigratingVMs to adjust loads

252 Fog Computing Xu et al [91] proposed a methodcalled ldquoDRAMrdquo to dynamically allocate resources in fogcomputing environment which can avoid both too highand too low loads e method first analyzes the loadbalance of different kinds of computing nodes And then itdesigns a fog resource allocation method to achieve loadbalance which allocates resources statically and migratesservices dynamically Oueis et al [92] studied the loadbalance problem in fog computing A custom fog clus-tering algorithm is proposed to solve the problem In theproblem several users need to offload computations andall of their demands need to be handled by local com-puting cluster

253 IoT Wang et al [93] established architecture of theenergy saving targeted system in industrial IoT In additionin order to predict sleep intervals they developed a sleepscheduling and a wake protocol which provide a better wayfor energy saving

3 Conclusion

is paper did a comprehensive study of QoS optimi-zation and energy saving in cloud computing edgecomputing fog computing and IoT models We sum-marized five main problems and analyzed their solutionsproposed by existing works By conducting this surveywe aim to help readers have a deeper understanding onthe concepts of different computing models and thetechniques of QoS optimization and energy saving inthese models

e investigated papers focus on issues about ensuringQoS and reducing SLA violations and resource manage-ment In the case of QoS assurance and SLA violationsreduction the main solution of QoS assurance is efficientVM management is solution can meet customersrsquo re-quirements through reasonable scheduling and integrationof VMs Most of resource management techniques are re-alized by reasonable scheduling of resources which canreduce the waste of VMs servers and traffic

Disclosure

is manuscript is an extension of A Survey of QoS Opti-mization and Energy Saving in Cloud Edge and IoT in e9th EAI International Conference on Cloud Computing

Conflicts of Interest

e authors declare that there are no conflicts of interestregarding the publication of this paper

Acknowledgments

is work was supported by the National Natural ScienceFoundation of China (Grant no 61702274) Natural ScienceFoundation of Jiangsu Province (Grant no BK20170958)and PAPD

Table 6 Work summary of load balancing in cloud computing

Problems Solutions Literatures AdvantagesServermanagement

An operation model that can balance cloudcomputing load and expand application [86] Saves energy by managing the number of servers running in

the system

Workloadmanagement

A hybrid approach [87] Reduces the footprint of carbon and allocates workload acrosscloud computing

An algorithm to dynamically manage theload [88] Manages the load evens the load distribution between

servers and allocates tasks between VMs

VMmanagement

A green power administration mechanism [89] Monitors and jointly allocates VM resources

An optimization system [90]Migrates VMs to adjust high and low loads without

interrupting services and balances the load of VMs runningon multiple physical machines

12 Complexity

References

[1] Q Zhang L Cheng and R Boutaba ldquoCloud computingstate-of-the-art and research challengesrdquo Journal of InternetServices and Applications vol 1 no 1 pp 7ndash18 2010

[2] W Shi J Cao Q Zhang Y Li and L Xu ldquoEdge computingvision and challengesrdquo IEEE Internet of Aings Journal vol 3no 5 pp 637ndash646 2016

[3] M Iorga L Feldman R Barton M J Martin N S Gorenand C Mahmoudi ldquoFog computing conceptual modelrdquo TechRep Recommendations of the National Institute of Standardsand Technology Gaithersburg MD USA 2018

[4] Nebbiolo ldquoFog vs edge computingrdquo Tech Rep NebbioloTechnologies Inc Milpitas CA USA 2018

[5] C T Do N H Tran C Pham M G R Alam J H Son andC S Hong ldquoA proximal algorithm for joint resource allo-cation andminimizing carbon footprint in geo-distributed fogcomputingrdquo in Proceedings of the 2015 International Con-ference on Information Networking (ICOIN) pp 324ndash329IEEE Siem Reap Cambodia January 2015

[6] S Vashi J Ram J Modi S Verma and C Prakash ldquoInternetof things (IoT) a vision architectural elements and securityissuesrdquo in Proceedings of the 2017 International Conference onI-SMAC (IoT in Social Mobile Analytics and Cloud)(I-SMAC) pp 492ndash496 IEEE Coimbatore India February2017

[7] M Mazzucco D Dyachuk and R Deters ldquoMaximizing cloudprovidersrsquo revenues via energy aware allocation policiesrdquo inProceedings of the 2010 IEEE 3rd International Conference onCloud Computing pp 131ndash138 IEEE Miami FL USA July2010

[8] Q He R Zhou X Zhang et al ldquoKeyword search for buildingservice-based systemsrdquo IEEE Transactions on Software En-gineering vol 43 no 7 pp 658ndash674 2016

[9] L Sun H Dong O K Hussain F K Hussain and A X LiuldquoA framework of cloud service selection with criteria inter-actionsrdquo Future Generation Computer Systems vol 94pp 749ndash764 2019

[10] M Mazzucco and D Dyachuk ldquoOptimizing cloud providersrevenues via energy efficient server allocationrdquo SustainableComputing Informatics and Systems vol 2 no 1 pp 1ndash122012

[11] Q He J Han F Chen et al ldquoQos-aware service selection forcustomisable multi-tenant service-based systems maturityand approachesrdquo in Proceedings of the 2015 IEEE 8th Inter-national Conference on Cloud Computing pp 237ndash244 IEEENew York NY USA July 2015

[12] L Sun J Ma H Wang Y Zhang and J Yong ldquoCloud servicedescription model an extension of usdl for cloud servicesrdquoIEEE Transactions on Services Computing vol 11 no 2pp 354ndash368 2015

[13] Y Wang Q He D Ye and Y Yang ldquoFormulating criticality-based cost-effective fault tolerance strategies for multi-tenantservice-based systemsrdquo IEEE Transactions on Software En-gineering vol 44 no 3 pp 291ndash307 2017

[14] S Mustafa K Bilal S U R Malik and S A Madani ldquoSla-aware energy efficient resource management for cloud en-vironmentsrdquo IEEE Access vol 6 pp 15004ndash15020 2018

[15] J Bi H Yuan M Tie and W Tan ldquoSla-based optimisation ofvirtualised resource for multi-tier web applications in clouddata centresrdquo Enterprise Information Systems vol 9 no 7pp 743ndash767 2015

[16] S Singh I Chana and R Buyya ldquoStar sla-aware autonomicmanagement of cloud resourcesrdquo IEEE Transactions on CloudComputing p 1 2017

[17] A Beloglazov and R Buyya ldquoEnergy efficient resourcemanagement in virtualized cloud data centersrdquo in Proceedingsof the 2010 10th IEEEACM International Conference onCluster Cloud and Grid Computing pp 826ndash831 IEEEComputer Society Melbourne Australia May 2010

[18] M Guazzone C Anglano and M Canonico ldquoEnergy-effi-cient resource management for cloud computing infra-structuresrdquo in Proceedings of the 2011 IEEE AirdInternational Conference on Cloud Computing Technology andScience pp 424ndash431 IEEE Athens Greece November 2011

[19] Y Sun J White and S Eade ldquoA model-based system toautomate cloud resource allocation and optimizationrdquo inProceedings of the International Conference on Model DrivenEngineering Languages and Systems pp 18ndash34 SpringerValencia Spain October 2014

[20] G Siddesh and K Srinivasa ldquoSla-driven dynamic resourceallocation on cloudsrdquo in Proceedings of the InternationalConference on Advanced Computing Networking and Securitypp 9ndash18 Springer Surathkal India December 2011

[21] S K Garg S K Gopalaiyengar and R Buyya ldquoSla-basedresource provisioning for heterogeneous workloads in avirtualized cloud datacenterrdquo in Proceedings of the Interna-tional Conference on Algorithms and Architectures for ParallelProcessing pp 371ndash384 Springer Melbourne AustraliaOctober 2011

[22] J Bi Z Zhu and H Yuan ldquoSla-aware dynamic resourceprovisioning for profit maximization in shared cloud datacentersrdquo in Proceedings of the International Conference onHigh Performance Networking Computing and Communi-cation Systems pp 366ndash372 Springer Singapore May 2011

[23] L Qi Y Chen Y Yuan S Fu X Zhang and X Xu ldquoA Qos-aware virtual machine scheduling method for energy con-servation in cloud-based cyber-physical systemsrdquo WorldWide Web vol 4 no 3 pp 1ndash23 2019

[24] A Beloglazov J Abawajy and R Buyya ldquoEnergy-aware re-source allocation heuristics for efficient management of datacenters for cloud computingrdquo Future Generation ComputerSystems vol 28 no 5 pp 755ndash768 2012

[25] L Qi R Wang C Hu S Li Q He and X Xu ldquoTime-awaredistributed service recommendation with privacy-preserva-tionrdquo Information Sciences vol 480 pp 354ndash364 2019

[26] Z Zhai B Cheng Y Tian J Chen L Zhao and M Niu ldquoAdata-driven service creation approach for end-usersrdquo IEEEAccess vol 4 pp 9923ndash9940 2016

[27] L Gu D Zeng S Guo A Barnawi and Y Xiang ldquoCostefficient resource management in fog computing supportedmedical cyber-physical systemrdquo IEEE Transactions onEmerging Topics in Computing vol 5 no 1 pp 108ndash119 2015

[28] L Ni J Zhang C Jiang C Yan and K Yu ldquoResource al-location strategy in fog computing based on priced timed petrinetsrdquo Ieee Internet of Aings Journal vol 4 no 5pp 1216ndash1228 2017

[29] L Wei T Yang F C Delicato et al ldquoOn enabling sustainableedge computing with renewable energy resourcesrdquo IEEECommunications Magazine vol 56 no 5 pp 94ndash101 2018

[30] P Lai Q He G Cui et al ldquoEdge user allocation with dynamicquality of servicerdquo in Proceedings of the International Con-ference on Service-Oriented Computing pp 86ndash101 SpringerToulouse France October 2019

[31] X Xu Q Liu X Zhang J Zhang L Qi and W Dou ldquoAblockchain-powered crowdsourcing method with privacy

Complexity 13

preservation in mobile environmentrdquo IEEE Transactions onComputational Social Systems vol 6 no 6 pp 1407ndash1419 2019

[32] O Rolik E Zharikov and S Telenyk ldquoMicrocloud-basedarchitecture of management system for IoT infrastructuresrdquoin Proceedings of the 2016 Aird International Scientific-Practical Conference Problems of Infocommunications Scienceand Technology (PIC SampT) pp 149ndash151 IEEE KharkivUkaraine October 2016

[33] W He S Guo Y Liang R Ma X Qiu and L Shi ldquoQos-awareand resource-efficient dynamic slicingmechanism for internetof thingsrdquo Computers Materials amp Continua vol 61 no 3pp 1345ndash1364 2019

[34] J Yao and N Ansari ldquoQos-aware fog resource provisioningand mobile device power control in IoT networksrdquo IEEETransactions on Network and Service Management vol 16no 1 pp 167ndash175 2018

[35] W Wang Y Jiang and W Wu ldquoMultiagent-based resourceallocation for energy minimization in cloud computing sys-temsrdquo IEEE Transactions on Systems Man and CyberneticsSystems vol 47 no 2 pp 1ndash16 2016

[36] K Krishnajyothi ldquoParallel data processing for effective dy-namic resource allocation in the cloudrdquo International Journalof Computer Applications vol 70 no 22 pp 1ndash4 2013

[37] M M Hassan B Song M S Hossain and A Alamri ldquoEf-ficient resource scheduling for big data processing in cloudplatformrdquo in Proceedings of the International Conference onInternet and Distributed Computing Systems pp 51ndash63Springer Calabria Italy September 2014

[38] C-M Wu R-S Chang and H-Y Chan ldquoA green energy-efficient scheduling algorithm using the dvfs technique forcloud datacentersrdquo Future Generation Computer Systemsvol 37 no 7 pp 141ndash147 2014

[39] Z Wang and X Su ldquoDynamically hierarchical resource-al-location algorithm in cloud computing environmentrdquo AeJournal of Supercomputing vol 71 no 7 pp 2748ndash2766 2015

[40] W-Y Lin G-Y Lin and H-Y Wei ldquoDynamic auctionmechanism for cloud resource allocationrdquo in Proceedings ofthe 2010 10th IEEEACM International Conference on ClusterCloud and Grid Computing pp 591-592 IEEE ComputerSociety Victoria Australia May 2010

[41] Y O Yazir C Matthews R Farahbod et al ldquoDynamic re-source allocation in computing clouds using distributedmultiple criteria decision analysisrdquo in Proceedings of the 2010IEEE 3rd International Conference on Cloud Computingpp 91ndash98 IEEE Washington DC USA July 2010

[42] W Lin J Z Wang C Liang and D Qi ldquoA threshold-baseddynamic resource allocation scheme for cloud computingrdquoProcedia Engineering vol 23 no 5 pp 695ndash703 2011

[43] X Xu S Fu L Qi et al ldquoAn IoT-oriented data placementmethod with privacy preservation in cloud environmentrdquoJournal of Network and Computer Applications vol 124pp 148ndash157 2018

[44] J Bokyun P Md Jalil L Daeho and S Doug Young ldquoEf-ficient computation offloading in mobile cloud computing forvideo streaming over 5 grdquo Computers Materials amp Continuavol 61 no 2 pp 439ndash463 2019

[45] H Sarbazi-Azad and A Y Zomaya ldquoEnergy-efficient resourceutilization in cloud computingrdquo in Large Scale Network-Centric Distributed Systems pp 377ndash408 Wiley-IEEE PressHoboken NJ USA 2014

[46] C-H Hsu K D Slagter S-C Chen and Y-C ChungldquoOptimizing energy consumption with task consolidation incloudsrdquo Information Sciences vol 258 no 3 pp 452ndash4622014

[47] C-H Hsu S-C Chen C-C Lee et al ldquoEnergy-aware taskconsolidation technique for cloud computingrdquo in Proceedingsof the 2011 IEEE Aird International Conference on CloudComputing Technology and Science pp 115ndash121 IEEE Ath-ens Greece 2011

[48] S K Panda and P K Jana ldquoAn efficient task consolidationalgorithm for cloud computing systemsrdquo in Proceedings of theInternational Conference on Distributed Computing and In-ternet Technology pp 61ndash74 Springer Bhubaneswar IndiaJanuary 2016

[49] L Yin J Luo and H Luo ldquoTasks scheduling and resourceallocation in fog computing based on containers for smartmanufacturingrdquo IEEE Transactions on Industrial Informaticsvol 14 no 10 pp 4712ndash4721 2018

[50] M Aazam and E-N Huh ldquoDynamic resource provisioningthrough fog micro datacenterrdquo in Proceedings of the 2015IEEE International Conference on Pervasive Computing andCommunication Workshops (PerCom Workshops) pp 105ndash110 IEEE St Louis MO USA March 2015

[51] B Jia H Hu Y Zeng T Xu and Y Yang ldquoDouble-matchingresource allocation strategy in fog computing networks basedon cost efficiencyrdquo Journal of Communications and Networksvol 20 no 3 pp 237ndash246 2018

[52] H Zhang Y Xiao S Bu D Niyato F R Yu and Z HanldquoComputing resource allocation in three-tier IoT fog net-works a joint optimization approach combining stackelberggame and matchingrdquo IEEE Internet of Aings Journal vol 4no 5 pp 1204ndash1215 2017

[53] J Tan T-H Chang and T Q Quelc ldquoMinimum energyresource allocation in fog radio access network with fronthauland latency constraintsrdquo in Proceedings of the 2018 IEEE 19thInternational Workshop on Signal Processing Advances inWireless Communications (SPAWC) pp 1ndash5 IEEE KalamataGreece June 2018

[54] D R de Vasconcelos R M de Castro Andrade andJ N de Souza ldquoSmart shadowndashan autonomous availabilitycomputation resource allocation platform for internet ofthings in the fog computing environmentrdquo in Proceedings of the2015 International Conference on Distributed Computing inSensor Systems pp 216-217 IEEE Fortaleza Brazil June 2015

[55] M Aazam M St-Hilaire C-H Lung and I LambadarisldquoPre-fog IoT trace based probabilistic resource estimation atfogrdquo in Proceedings of the 2016 13th IEEE Annual ConsumerCommunications amp Networking Conference (CCNC) pp 12ndash17 IEEE Las Vegas NV USA January 2016

[56] N D Tung L L Bao and B Vijay ldquoPrice-based resourceallocation for edge computing a market equilibrium ap-proachrdquo IEEE Transactions on Cloud Computing p 1 2018

[57] X Xu C He Z Xu L Qi S Wan and M Z A BhuiyanldquoJoint optimization of offloading utility and privacy for edgecomputing enabled IoTrdquo IEEE Internet of Aings Journal2019

[58] X Xu Y Chen X Zhang Q Liu X Liu and L Qi ldquoAblockchain-based computation offloading method for edgecomputing in 5G networksrdquo Software Practice and Experi-ence 2019

[59] X Xu Y Li T Huang et al ldquoAn energy-aware computationoffloading method for smart edge computing in wirelessmetropolitan area networksrdquo Journal of Network and Com-puter Applications vol 133 pp 75ndash85 2019

[60] X Xu Y Xue L Qi et al ldquoAn edge computing-enabledcomputation offloading method with privacy preservation forinternet of connected vehiclesrdquo Future Generation ComputerSystems vol 96 pp 89ndash100 2019

14 Complexity

[61] G Yeting L Fang X Nong and C Zhengguo ldquoTask-basedresource allocation bid in edge computing micro datacenterrdquoComputers Materials amp Continua vol 61 no 2 pp 777ndash7922019

[62] X Chen L Jiao W Li and X Fu ldquoEfficient multi-usercomputation offloading for mobile-edge cloud computingrdquoIEEEACM Transactions on Networking vol 24 no 5pp 2795ndash2808 2016

[63] B Gao L He X Lu C Chang K Li and K Li ldquoDevelopingenergy-aware task allocation schemes in cloud-assisted mo-bile workflowsrdquo in Proceedings of the 2015 IEEE InternationalConference on Computer and Information Technology Ubiq-uitous Computing and Communications Dependable Auto-nomic and Secure Computing Pervasive Intelligence andComputing pp 1266ndash1273 IEEE Liverpool UK October2015

[64] X Xu X Zhang H Gao Y Xue L Qi andW Dou ldquoBecomeblockchain-enabled computation offloading for IoT in mobileedge computingrdquo IEEE Transactions on Industrial Infor-matics 2019

[65] W Yifei W Zhaoying G Da and Y F Richard ldquoDeepq-learning based computation offloading strategy for mobileedge computingrdquo Computers Materials amp Continua vol 59no 1 pp 89ndash104 2019

[66] M Barcelo A Correa J Llorca A M Tulino J L Vicarioand A Morell ldquoIoT-cloud service optimization in nextgeneration smart environmentsrdquo IEEE Journal on SelectedAreas in Communications vol 34 no 12 pp 4077ndash40902016

[67] B Cheng M Wang S Zhao Z Zhai D Zhu and J ChenldquoSituation-aware dynamic service coordination in an IoTenvironmentrdquo IEEEACM Transactions On Networkingvol 25 no 4 pp 2082ndash2095 2017

[68] V Angelakis I Avgouleas N Pappas and D Yuan ldquoFlexibleallocation of heterogeneous resources to services on an IoTdevicerdquo in Proceedings of the 2015 IEEE Conference onComputer Communications Workshops (INFOCOMWKSHPS) pp 99-100 IEEE Hong Kong China April 2015

[69] S Li Q Ni Y Sun G Min and S Al-Rubaye ldquoEnergy-ef-ficient resource allocation for industrial cyber-physical IoTsystems in 5g erardquo IEEE Transactions on Industrial Infor-matics vol 14 no 6 pp 2618ndash2628 2018

[70] X Liu Z Qin Y Gao and J A McCann ldquoResource allocationin wireless powered IoT networksrdquo IEEE Internet of AingsJournal vol 6 no 3 pp 4935ndash4945 2019

[71] W Ejaz and M Ibnkahla ldquoMulti-band spectrum sensing andresource allocation for IoT in cognitive 5g networksrdquo IEEEInternet of Aings Journal vol 5 no 1 pp 150ndash163 2017

[72] G Colistra V Pilloni and L Atzori ldquoTask allocation in groupof nodes in the IoT a consensus approachrdquo in Proceedings ofthe 2014 IEEE International Conference on Communications(ICC) pp 3848ndash3853 IEEE Sydney Australia June 2014

[73] J Li Q Sun and G Fan ldquoResource allocation for multiclassservice in IoTuplink communicationsrdquo inProceedings of the 20163rd International Conference on Systems and Informatics (ICSAI)pp 777ndash781 IEEE Shanghai China November 2016

[74] L I Zheng and K H Liu ldquoDynamic bandwidth resourceallocation algorithm in internet of things and its applicationrdquoComputer Engineering vol 38 no 17 pp 16ndash19 2012

[75] K Gai and M Qiu ldquoOptimal resource allocation using re-inforcement learning for IoT content-centric servicesrdquo Ap-plied Soft Computing vol 70 pp 12ndash21 2018

[76] X Xu W Dou X Zhang and J Chen ldquoEnreal an energy-aware resource allocation method for scientific workflow

executions in cloud environmentrdquo IEEE Transactions onCloud Computing vol 4 no 2 pp 166ndash179 2016

[77] K Bousselmi Z Brahmi and M M Gammoudi ldquoEnergyefficient partitioning and scheduling approach for scientificworkflows in the cloudrdquo in Proceedings of the 2016 IEEEInternational Conference on Services Computing (SCC)pp 146ndash154 IEEE San Francisco CA USA June 2016

[78] Y Sonia C Rachid K Hubert and G Bertrand ldquoMulti-objective approach for energy-aware workflow scheduling incloud computing environmentsrdquo Ae Scientific World Jour-nal vol 2013 Article ID 350934 13 pages 2013

[79] F Cao Efficient Scientific Workflow Scheduling in CloudEnvironment Southern Illinois University Carbondale ILUSA 2014

[80] Z Li J Ge H Hu W Song H Hu and B Luo ldquoCost andenergy aware scheduling algorithm for scientific workflowswith deadline constraint in cloudsrdquo IEEE Transactions onServices Computing vol 11 no 4 pp 713ndash726 2015

[81] M Khaleel and M M Zhu ldquoEnergy-aware job managementapproaches for workflow in cloudrdquo in Proceedings of the 2015IEEE International Conference on Cluster Computingpp 506-507 IEEE Chicago IL USA September 2015

[82] J Shi J Luo F Dong and J Zhang ldquoA budget and deadlineaware scientific workflow resource provisioning and sched-uling mechanism for cloudrdquo in Proceedings of the 2014 IEEE18th International Conference on Computer Supported Co-operative Work in Design (CSCWD) pp 672ndash677 IEEEHsinchu Taiwan January 2014

[83] Y Ge Y Zhang Q Qiu and Y-H Lu ldquoA game theoreticresource allocation for overall energy minimization in mobilecloud computing systemrdquo in Proceedings of the 2012ACMIEEE International Symposium on Low Power Electronicsand Design pp 279ndash284 ACM Redondo Beach CA USAAugust 2012

[84] X Wang Y Wang and C Yue ldquoAn energy-aware bi-leveloptimization model for multi-job scheduling problems undercloud computingrdquo Soft Computing vol 20 no 1 pp 303ndash3172016

[85] R Yanggratoke F Wuhib and R Stadler ldquoGossip-basedresource allocation for green computing in large cloudsrdquo inProceedings of the 2011 7th International Conference onNetwork and Service Management pp 1ndash9 IEEE ParisFrance October 2011

[86] A Paya and D C Marinescu ldquoEnergy-aware load balancingand application scaling for the cloud ecosystemrdquo IEEETransactions on Cloud Computing vol 5 no 1 pp 15ndash272015

[87] V D Justafort R Beaubrun and S Pierre ldquoA hybrid ap-proach for optimizing carbon footprint in intercloud envi-ronmentrdquo IEEE Transactions on Services Computing vol 12no 2 pp 186ndash198 2016

[88] R Panwar and B Mallick ldquoLoad balancing in cloud com-puting using dynamic load management algorithmrdquo inProceedings of the 2015 International Conference on GreenComputing and Internet of Aings (ICGCIoT) pp 773ndash778IEEE Noida India October 2015

[89] C-T Yang K-C Wang H-Y Cheng C-T Kuo andW C C Chu ldquoGreen power management with dynamicresource allocation for cloud virtual machinesrdquo in Proceedingsof the 2011 IEEE International Conference on High Perfor-mance Computing and Communications pp 726ndash733 IEEEAlberta Canada September 2011

[90] C-T Yang H-Y Cheng and K-L Huang ldquoA dynamicresource allocation model for virtual machine management

Complexity 15

on cloudrdquo in Proceedings of the International Conference onGrid and Distributed Computing pp 581ndash590 SpringerGangneug Korea December 2011

[91] X Xu F Shucun C Qing et al ldquoDynamic resource allocationfor load balancing in fog environmentrdquo Wireless Commu-nications amp Mobile Computing vol 2018 no 2 Article ID6421607 15 pages 2018

[92] J Oueis E C Strinati and S Barbarossa ldquoe fog balancingload distribution for small cell cloud computingrdquo in Pro-ceedings of the 2015 IEEE 81st Vehicular Technology Confer-ence (VTC Spring) pp 1ndash6 IEEE Glasgow UK May 2015

[93] K Wang Y Wang Y Sun S Guo and J Wu ldquoGreen in-dustrial internet of things architecture an energy-efficientperspectiverdquo IEEE Communications Magazine vol 54 no 12pp 48ndash54 2016

16 Complexity

environment and reduce SLA violations So the paymentefficiency of cloud services can be improved Beloglazov andBuyya [17] proposed a system to manage energy in the clouddata center By continuously integrating VMs and dy-namically redistributing VMs the system can achieve thegoal of saving energy and providing a high QoS level at thesame time Guazzone et al [18] proposed an automaticmanagement system (see Figure 4) for resources to providecertain QoS levels and reduce energy consumption Re-source manager of the framework in Figure 4 combinesvirtualization technologies and control-theoretic technolo-gies Virtualization technologies deploy each application toindependent VM And control-theoretic technologies realizethe automatic management of computer performance andenergy consumption In addition the resource managerconsists of several independent components named Ap-plication Manager Physical Machine Manager and Mi-gration Manager Different from traditional static methodsthis method can both fit the changing workloads dynami-cally and achieve remarkable results in reducing QoS vio-lations Sun et al [19] established a model to simplify thedecision of cloud resource allocation and realize the inde-pendent allocation of resources e optimal resourceconfiguration can be obtained so the QoS requirements canbe well met Siddesh and Srinivasa [20] explored theproblems of dynamic resource allocation and SLA assuranceey proposed a framework to deal with heterogeneousworkload types by dynamically planning computing capacityand assessing riskse framework uses scheduling methodsto reduce SLA violation risks and maximize revenues inresource allocation

Garg et al [21] proposed a resource allocation strategyfor VM dynamic allocation e strategy can improve re-source utilization increase providersrsquo profits and reduceSLA violations Jing et al [22] proposed a new dynamicallocating technique using the mixed queue model meetingcustomersrsquo different requirements of performance by pro-viding virtualized resources to each layer of virtualizedapplication services All these methods can reasonablyconfigure resources in the cloud data center improve systemperformance reduce additional costs of using resources andmeet the required QoS

Qi et al [23] proposed a QoS-aware VM schedulingstrategy named QVMS to satisfy QoS Firstly the schedulingproblem is transformed into a problem with several ob-jectives And then the optimal VM migration method isfound according to the genetic algorithm e schedulingstrategy can effectively manage resources in the networkphysical system thus reducing the energy consumption andimproving QoS levels Qi et al [25] proposed a servicerecommendation strategy by considering the time factor toimprove the traditional location-sensitive hash technologye policy emphasizes the influence of dynamic factors onQoS and the protection of user privacy

Table 1 shows a summary of the abovementioned workse solution of the problems in Table 1 can improve QoS incloud computing environment Server management refers todynamically allocating powering serversrsquo switches Work-loads consolidation refers to combining work to save energy

VM management refers to reasonable scheduling or inte-gration of VMs to achieve better performance Self-man-agement refers to the realization of self-management ofresources which can achieve higher efficiency Resourcemanagement refers to the correct allocation of resources toreduce waste Service management is about making rea-sonable service choices [26]

212 Fog Computing Gu et al [27] used fog computing toprocess a large amount of data generated by medical devicesand built Fog Computing Supported Medical Cyber-Phys-ical System (FC-MCPS) In order to reduce the cost of FC-MCPS research studies were carried out on the joint of basestation task assignment and VM layout e problem ismodeled as a mixed integer linear programming (MILP) Atwo-stage heuristic algorithm based on linear programming(LP) is proposed to solve the problem Ni et al [28] proposeda resource allocation approach based on fog computingwhich enables users to select resources independently Inaddition this approach takes into account the price and timerequired to finish the job Formula (4) [28] is used to definethe credibility BCreij of Resource Rj received from useriwhen the user interacts with Resource Ri

BCreij ω1λresp + ω2cexec + ω3 1 minus ηreboot( 1113857 + ω4μrel

(4)

where the value of ωkϵ[0 1] 11139364k1 ωk 1 which can be

determined by the user or the actual situation λresp cexecηreboot and μrel are the response speed of the corre-sponding index service the efficiency of execution the speedof restart and the reliability respectively

213 Edge Computing Wei et al [29] proposed a unifiedframework in the sustainable edge computing to save energyincluding the energy that is distributed and renewable Andthe architecture can combine the system that supply energyand edge services which can make full use of renewableenergy and provide better QoS Lai et al [30] proposed anoptimized allocation method for edge usersemethod cannot only maximize the amount of resources allocated tousers but also consider the dynamic QoS level of users Soedge user allocation problem can be made more general andimproving the quality of experience

214 MEC Xu et al [31] used block chain to improve thetraditional crowdsourcing technology Firstly they proposeda mobile crowdsourcing framework using block chaintechnology to protect user privacy en they used dynamicprogramming strategy of clustering algorithm to classifyrequesters Finally they generated service policies to balanceprofits and energy consumption

215 IoT Rolik et al [32] proposed a method to build aframework of IoT infrastructure based on microcloud emethod can help users use resources rationally reduce thecost of managing infrastructure and improve the quality of

4 Complexity

life of consumers He et al [33] proposed a dynamic networkslice strategy e network slice can be dynamically adjustedaccording to the time-varying resource demands is methodcan improve the utilization of the underlying resources andbetter meet different QoS demands Yao and Ansari [34]proposed an algorithm to determine the number of VMs to berented and to control the power supply us the cost of thesystem can be minimized and the QoS can be improvedFormula (5) [34] is used to limit the delay requirement of QoSe total delay must not exceed the computation deadline ofeach task and the total delay is composed of wireless trans-mission delay and fog processing delay

tci + t

wi leDi foralliϵN (5)

where c and w respectively represents fog processing andwireless transmission i denotes a location tc

i represents thedelay of processing tw

i represents the delay of wirelesstransmission Di denotes the deadline and N denotes dif-ferent locations

22 Resource Management and Allocation Rational alloca-tion of resources is an effective means to save energy

221 Cloud Computing Wang et al [35] introduced anallocation method based on distributed multiagent to al-locate VMs to physical machines e method can realize

VM consolidation and consider the migration costs si-multaneously In addition a VM migration mechanismbased on local negotiation is proposed to avoid unnecessaryVM migration costs Hassan et al [37] established a for-mulation of universal problem and proposed a heuristicalgorithm which has optimal parameters Under this for-mulation dynamic resource allocation can be made to meetthe QoS requirements of applications And the cost neededfor dynamic resource allocation can be minimized with thisalgorithm Wu et al [38] proposed a scheduling algorithmbased on the technology that can scale the voltage frequencydynamically in cloud computing e algorithm can allocateresources for performing tasks and realize low powerconsumption network infrastructure Compared with otherschemes this scheme not only sacrifices the performance ofexecution operations but also saves more energy

Sarbazi and Zomaya [45] used two job consolidationheuristic methods to save energy One is MaxUtil to betterutilize resources and the other is Energy-Conscious TaskConsolidation to reduce energy consumption ese twomethods can promote the concurrent execution of multipletasks and improve the energy efficiency Hsu et al [46]proposed a job consolidation technique to minimize energyconsumption Formula (6) [46] defines the energy con-sumption of VM Vi from time t0 to tm in the cluster isdefined And formula (7) [46] defines the total energyconsumption in a virtual cluster VCk in the period Inaddition the proposed technique limits the CPU usage and

Reference machine

Reference machine

SLA

SLA

Application A1

Application An

TIER 1

TIER 2

TIER 3

TIER 1

TIER 2

TIER 3

Cloud infrastructure

VM

VM

VM

VM

VM

VM

Performance measuresEnergy consumptions

Resource costs

Resourcemanager

Migrationmanager

Physicalmachinemanagers

Applicationmanagers

Figure 4 Framework of the proposed three-fold system [18]

Complexity 5

Tabl

e1

Worksummaryof

QoS

guaranteeing

orSL

Aassurancein

clou

dcompu

ting

Subp

roblem

sSolutio

nsLiteratures

Advantages

Server

managem

ent

Apo

licyof

dynamicallocatio

nof

poweringserverss

witches

[71

0]Maxim

izes

benefitsim

proves

QoSa

ndminim

izes

power

consum

ption

Workloads

consolidation

Atechniqu

eforconsolidatingworkloads

[14]

Achievesenergy

saving

sby

usingfewestserverswhile

redu

cing

SLA

violations

VM

managem

ent

Anarchitecturethat

canadministrate

itselfa

ndamixed

queuingmod

el[15]

Makes

morerevenu

esandmeets

different

requ

irem

ents

ofcustom

ersanddecidesthe

numberof

VMsforeach

layerof

avirtuala

pplication

Asystem

tointegrateanddynamically

redistribu

teVMs

[17]

Achievesthe

goalofsaving

energy

throughVM

integrationandprovidesah

ighQoS

levelat

thesametim

e

AQoS-awareVM

schedu

lingstrategy

named

QVMS

[23]

Effectiv

elymanages

resourcesin

thenetworkph

ysical

system

toredu

cetheenergy

consum

ptionandim

proves

QoS

AVM

integrationmetho

dwith

severaltargets

[24]

Savesenergy

andredu

cesSL

Aviolations

byapplying

different

strategies

todifferent

load

states

oftheho

st

Self-managem

ent

Atechno

logy

named

STAR

[16]

Redu

cesSL

Aviolations

andim

proves

paym

ente

fficiency

ofclou

dservices

Adynamic

resource

managem

entsystem

[18]

Self-manages

theresourceso

fcloud

infrastructurestoprovidea

ppropriateQoS

andfitsthe

changing

workloads

dynamically

Resource

managem

ent

Amod

elthat

canrealizetheindepend

entallocatio

nof

resources

[19]

Obtains

theop

timal

resource

confi

guratio

nmeets

theQoS

requ

irem

entsand

provides

econ

omical

clou

dresources

Adynamic

resource

allocatio

nstrategy

[20

21]

Redu

cesSL

Aandmaxim

izes

revenu

esandresource

utilizatio

non

theclou

d

Amixed

queuemod

el[22]

Reason

ably

confi

gurestheresourcesin

theclou

ddata

centerimproves

thesystem

performancereduces

theadditio

nalcosto

fusin

gresourcesmeetstherequ

ired

QoSand

provides

virtualresou

rces

toeach

layerof

virtuala

pplicationservices

Service

managem

ent

Aun

ified

semantic

mod

elthat

candescribe

clou

dservice

[12]

Improves

theability

ofmod

elon

servicerank

ingandenriches

thelang

uage

expressio

n

Arecommendatio

nservicestrategy

[25]

Emph

asizes

theinflu

ence

oftim

efactorson

QoS

andim

proves

thetradition

allocatio

n-sensitive

hash

techno

logy

toprotectusersrsquoprivacy

Aclou

dserviceselectionmetho

dusingfuzzymeasure

and

Cho

quet

integral

andafram

eworkbasedon

priority

[9]

Selectsservicewhenhistorical

inform

ationisinsufficientto

determ

inethecriteria

relatio

nships

andweigh

ts

reeserviceselectionmetho

dsthat

supp

ortQ

oSandcan

combine

multitenants

ervice-based

system

s[11]

Achievesthree

degreeso

fmultitenantm

aturity

which

ismoreeffi

cientthanthetradition

alsin

gle-user

approach

Afault-tolerant

strategy

basedon

multitenants

ervice

criticality

[13]

Guaranteesthequ

ality

ofthemultitenant-basedservicesystem

Aservice-basedsystem

supp

ortin

gkeyw

ordsearch

[8]

Effectiv

elyhelpssystem

engineerswho

areno

tfam

iliar

with

service-oriented

architecture

techno

logy

tobu

ildservice-oriented

system

s

6 Complexity

merges tasks in virtual clusters Once a task migrationhappens the energy cost model will take into account thenetwork latency Sarbazi-Azad and Zomaya [45] and Hsuet al [46] both maximize the benefit of cloud resources byusing task merging techniques Sarbazi-Azad and Zomaya[45] uses a greedy algorithm called MaxUtil While Hsuet al [46] takes into account the network latency associatedwith task migration So in [46] a 17 improvement isachieved over MaxUtil

E0m Vi( 1113857 1113944m

t0Et Vi( 1113857 (6)

E0m VCk( 1113857 1113944n

i0E0m Vi( 1113857 (7)

where Et is the energy consumption in unit time and n is thenumber of VMs in the cluster

Hsu et al [47] proposed a task integration technologybased on the energy perception According to the charac-teristics of most cloud systems the principle of using 70CPU is proposed to administrate job integration amongvirtual clusters is technology is very effective in reducingthe amount of energy consumed in cloud systems bymerging tasks Panda and Jana [48] proposed an algorithmwith several criteria to combine tasks e algorithm notonly considers the time needed for processing jobs but alsoconsiders the utilization rate of VMs And the algorithm ismore energy efficient because it takes into account not onlythe processing time but also the utilization rate of VMsWang and Su [39] proposed a resource allocation algorithmto deal with wide range of communication between nodes incloud environment is algorithm uses recognition tech-nology to dynamically distribute jobs and nodes accordingto computing ability and factors of storage And it can re-duce the traffic when allocating resources because it usesdynamic hierarchy Lin et al [40] proposed a dynamicauction approach for resource allocation e approach canensure that even if there are many users and resourcesproviders will have reasonable profits and computing re-sources will be allocated correctly Yazir et al [41] proposeda new method to manage resources dynamically and au-tonomously Firstly resource management is split into jobsand each job is executed by autonomous nodes Secondautonomous nodes use the method called PROMETHEE toconfigure resources Krishnajyothi [36] proposed a frame-work which can implement parallel task processing to solvethe problem of low efficiency when submitting large tasksCompared with the static framework this framework candynamically allocate VMs thus reducing costs and the timeof processing tasks Lin et al [42] proposed a method toallocate resources dynamically by using thresholds Becausethis method uses the threshold value it can optimize thereallocation of resources improve the usage of resourcesand reduce the cost Xu et al [43] proposed a data placementstrategy named IDP for the data generated by IoTs devices toachieve reasonable data placement In this way the privacyof these data can be protected while resources are allocatedreasonably Jo et al [44] proposed a computing offload

framework under 5G network e framework transfers thecomputing burden to the cloud thus reducing the com-puting load of clients and the communication cost

Table 2 shows a summary of the abovementioned workse problem of resource allocation and management incloud computing can be divided into problems in Table 2VM management is about a reasonable configuration ofVMs Resource allocation represents the dynamic andflexible allocation of resources Task integration refers tocombining tasks to save energy and improve efficiency

222 Fog Computing Yin et al [49] established a newmodel of scheduling jobs which applies containers In orderto make sure that jobs can be finished on time a jobscheduling algorithm is developed e algorithm can alsooptimize the number of tasks that can be performed togetheron the nodes in fog computing And this paper proposes aredistribution mechanism to shorten the delay of tasksese methods are very effective in reducing task delaysAazam and Huh [50] established a framework to admin-istrate resources effectively in the mode of fog computingConsidering that there are various types of objects anddevices the connection between them may be volatile So amethod for predicting and administrating resources isproposed e method considers that any object or devicecan quit using resources at anytime Cuong et al [5] studiedthe allocating resources jointly problem and the problem ofcarbon footprint minimization in fog data center Formula(8) [5] is used to denote the energy consumption of serversIn addition a distributed algorithm is proposed to solve theproblem of wide range optimization

P(y) C middot Pidle + Ppeak minus Pidle1113872 1113873 middot y middot κ (8)

where P(y) represents the power supply required by theservers in a data center y represents the video stream κdenotes a conversion factor that converts the video streaminto workload C represents the data centerrsquos load capacityand Pidle and Ppeak respectively represent the idle power andpeak power of the servers

Jia et al [51] studied the problem of computing resourceallocation in fog computing network with three levelsFirstly the problem of resource allocation is transformedinto a bilateral matching optimal problem And then a bi-matching approach is proposed for this problem which canimprove the performance of the system and obtain highercost efficiency Zhang et al [52] proposed a framework forjoint optimization under fog computing to allocate fognodesrsquo finite computing resources e framework canachieve the best allocation and effectively improve thenetworksrsquo performance Tan et al [53] presented a method toallocate computing and communication resources emethod transfers computing jobs to remote cloud and nodesand simplifies edge nodesrsquo computing and computing en-ergy Vasconcelos et al [54] developed a platform to allocateresources accessible to client devices in fog computingenvironment allocating the resources of devices near thehost to meet the applications needs for rapid response tocomputing resources Aazam et al [55] presented a method

Complexity 7

to estimate and manage resource in fog computing emethod is based on the fluctuation of customer abandon-ment probability type and price of service and so on

Table 3 shows a summary of the abovementioned workse problems in Table 3 are also derived from resourceallocation and management problems Task allocationrepresents the scheduling and redistribution of tasks Re-source allocation is still about the dynamic and flexibleallocation of resources Low latency refers to taking shorttime to configure and manage resources which can improveefficiency

223 Edge Computing Tung et al [56] proposed a newframework for resource allocation based on market needse resources come from edge nodes (ENs) with limitedheterogeneous capabilities and are allocated to multiplecompeting services on the network edge Generating amarket equilibrium solution by reasonably pricing ENs canobtain the maximum utilization of marginal computingresources Xu et al [57] proposed a strategy to optimizeoffloading and privacy protection is strategy shifts tasksfirstly to improve the resource utilization of resource-limited

edge cells And then it balances QoS performance andprivacy protection to achieve joint optimization

Xu et al [58] proposed an offload strategy for edgecomputing under 5G network which uses block chaintechnologye optimal strategy is further obtained by usingthe balanced offloading method It solves the problem ofdata loss under the condition of transmission delay which iscaused by the uneven requirements of user equipments onresources Xu et al [59] proposed a computational off-loading method named EACO to reduce the energy con-sumption in smart computing models Figure 5 showsarchitecture of smart edge computing where the shortestpath is used to unload tasks EACO uses genetic algorithmsto reduce the energy consumption for operating edgecomputing nodes and improve the efficiency of performingcomplex computing tasks Xu et al [60] proposed a com-putational offloading strategy for edge computing to protectthe privacy of interconnected vehicle networks ey firstlyanalyzed privacy conflicts of tasks And then they designedthe communication route to obtain routing vehicles whichcan achieve the optimization of several objectives Yetinget al [61] proposed a unique resource allocation mechanisme mechanism takes each individual task as the basis for

Table 2 Work summary of resource allocation and management in cloud computing

Subproblems Solutions Literatures Advantages

VMmanagement

A VM allocation method based on severaldistributed agents [35] Centralizes VMs to physical machines and reduces the

overall energy costA framework which can implement parallel

task processing [36] Dynamically allocates VMs for large tasks

Resourceallocation

A general problem formula and a heuristicalgorithm with optimization parameters [37]

Dynamically allocates resources according to QoSrequirements and realizes energy saving by optimizing the

number of serversA scheduling algorithm based on thetechnology that can scale the voltage

frequency dynamically[38] Ensures the performance of executing jobs while

implementing green computing

A fuzzy pattern recognition technology [39] Reduces the traffic when allocating resources and usesdynamic hierarchy

A dynamic auction approach [40]Guarantees profits of providers and allocates resourcescorrectly when there are a large amount of users and

resources

PROMETHEE [41]Allows node agents to decompose and execute tasksautonomously to improve the flexibility of resource

allocation

A method using thresholds [42]Optimizes resource reallocation improves resource usagereduces cost and studies resource allocation strategies at the

application level

IoT-oriented data placement (IDP) method [43] Protects data privacy allocates resources reasonably andfocuses on the placement method of IoT data

A computing offload framework under 5Gnetwork [44] Reduces the computing load of clients and the

communication cost

Taskconsolidation

Two job consolidation heuristics namedldquoMaxUtilrdquo and energy-conscious task

consolidation[45] Promotes concurrent execution of multiple tasks and

improves energy efficiency

A job consolidation technique aiming atenergy saving [46 47]

Reduces the power consumption of cloud system protectsdata privacy allocates resources reasonably limits CPU

usage and merges tasks in virtual clusters

An algorithm with several criteria [48]Takes into account the job processing time and the VM

utilization rate simultaneously Dramatically reduces energyconsumption compared with state-of-the-art works

8 Complexity

resource allocation rather than for the whole service It reducesthe packet loss rate and saves energy by unloading services

224 MEC Chen et al [62] studied the problem of com-puting unloading with several users in the environment ofMEC with wireless interference which have many channels Inaddition a distributed algorithm for computing unloading isdevelopede algorithm can perform the unloading well evenwhen there are a large number of users Gao et al [63] built aquadratic binary program which is able to assign tasks inmobile cloud computing environment Two algorithms arepresented to obtain the optimal solution Both of these heuristicalgorithms can effectively solve the task assignment problemXu et al [64] proposed an offloadingmethod using block chaintechnology It can guarantee the loss of data in offloading tasksunder edge computing And it can solve the problem of re-source requests out of proportion due to the limited load of

edge computing equipment during task transfer Yifei et al [65]proposed a model-free reinforcement learning framework tosolve the problem of computational unloadingis model canbe applied to the computational unloading with time-changingcomputing requests

225 IoT Barcelo et al [66] expressed the problem ofservice allocation [67] as a mixed flow problem with min-imum cost which can be solved by LP solving this serviceallocation problem can solve the problems of unbalancednetwork load and delay of end-to-end service And it canalso figure out the problem of excessive consumption ofelectricity brought by the architecture of centralized cloudAngelakis et al [68] assigned the requirements of servicesresources to heterogeneous network interfaces of equipmentsSo more heterogeneous network interfaces can be used by alarge amount of services

Mobile devices

Mobile devices ECN

ECN

ECNECN

AP

AP

AP

APAP

Figure 5 An architecture of smart edge computing [59]

Table 3 Work summary of resource allocation and management in fog computing

Problems Solutions Literatures AdvantagesTaskallocation

A new model of scheduling jobs and aredistribution mechanism [49] Finishes jobs on time optimizes the number of tasks and

shortens the delay of tasks

Resourceallocation

A framework to administrate resources and amethod to predict and administrate resources [50]

Assists service providers to predict the amount of availableresources based on different types of service customers anddeals with the phenomenon that objects or devices withdraw

from resource utilization at any time

A bi-matching approach [51] Improves the performance of the system and obtains highercost efficiency

A framework for joint optimization [52] Optimizes resource allocation and improves networkperformance

A method to allocate computing andcommunication resources [53]

Allocates computing and communication resources transferscomputing jobs to remote cloud and nodes simplifies edge

nodesrsquo computing and saves computing energyA platform to allocate resources accessible to

client devices [54] Enables rapid response to computing resources and allocatesthe resources of devices near the host

A framework to manage resources [55] Considers the fluctuation of the customer abandonmentprobability and service types

Low latency A distributed algorithm [5] Solves the problem of wide range optimization and allocatesresources jointly

Complexity 9

Li et al [69] proposed communication framework in 5Gand studied the problem of allocating power andchannels So the signal data in the channel can beavailable and the total energy efficiency can be maximumFormula (9) [69] shows how to calculate the energy ef-ficiency of a system

U 1113944K

k11113944

M

i1EES

iK + 1113944K

K11113944

N

i1EEA

jk (9)

where EESiK and EEA

jk respectively denote the energy ef-ficiency of sensor S and actuator A on channels e sets ofsensors actuators and channels are respectively repre-sented as s S1 S2 SM1113864 1113865 A A1 A2 AN1113864 1113865 andC C1 C2 CK1113864 1113865

Liu et al [70] studied the problem of allocating resourcesefficiently on IoT that supplies wireless power In thismethod users are first grouped into accessible channelsAnd then power distribution of users grouped in the samechannel is studied to improve throughput of the networkis method can allocate finite resources to a large group ofusers Ejaz and Ibnkahla [71] proposed the resource allo-cation framework with several bands under cognitive 5GIoT In the highly dynamic environment of the IoTmultiband method can manage resources more flexiblyand reduce more energy consumption In addition areconstruction approach with several levels is proposed toallocate resources reasonably for applications with dif-ferent needs of QoS Colistra et al [72] proposed aprotocol which is distributed and optimal to allocateresources in heterogeneous IoT Because this protocol hasexcellent adaptability when changing topology of net-work it can distribute resources evenly among nodes Jianet al [73] proposed a multilevel allocating resources al-gorithm for IoT communication using advanced tech-nology e algorithm uses hierarchical structure and hasfast data processing rate and very low latency in a satu-rated or not saturated environment

Zheng and Liu [74] proposed a new algorithm to allocatebandwidth dynamically for controlling remote computers inthe IoT is method can reduce the error of signal re-construction under the same bandwidth and make thebandwidth allocation of IoT more reasonable Gai and Qiu[75] used reinforcement learning mechanisms to allocateresources to achieve high Quality of Experience ismethod can effectively solve the resource allocation prob-lems caused by the mismatch of service quality and complexservice providing condition in the IoT

Table 4 shows a summary of the above works eproblem in Table 4 represents the realization of dynamic andflexible allocation of resources e resources here canrepresent channels bandwidth and power

23 Scientific Workflow Execution Implementing scientificworkflows especially in heterogeneous environments canreduce resource waste and reduce energy costs Scientificworkflow can be obtained by reasonably allocating resourcesand dynamically deploying VMs

231 Cloud Computing Xu et al [76] proposed a resourceallocation method called EnRealan to solve the problem ofenergy consumption e dynamic deployment of VMs isgenerally adopted to execute scientific workflows Bousselmiet al [77] proposed a scheduling method based on energyperception for executing scientific workflows in cloudcomputing At first an algorithm of splitting workflows forenergy minimization is presented which can achieve a highparallelism without huge energy consumption en aheuristic algorithm used to optimize cat swarm is proposedfor the created partitions e algorithm can minimize thetotal consumption of energy and the execution time ofworkflows Sonia et al [78] proposed a workflow schedulingmethod with several objects and hybrid particle swarmoptimization algorithm In addition a method for dy-namically scaling voltage and frequency is proposed emethod canmake the processors work at any voltage level soas to minimize the energy consumption in the process ofworkflow scheduling Both Bousselmi et al [77] and Soniaet al [78] use scheduling method to achieve scientificworkflows and study the problem of energy consumptione difference is that Bousselmi et al [77] focuses on in-tensive computing tasks while Sonia et al [78] focuses onworkflow scheduling on heterogeneous computing systems

Cao [79] established a scheduling algorithm of scientificworkflows with an objective of energy savingis algorithmcan enable service providers to gain high profits and reduceusersrsquo overhead at the same time Li et al [80] proposed ascheduling algorithm based on cloud computing which canminimize cost of performing workflows within a specifiedtime In addition the rented VM was modified to save costfurther Khaleel and Zhu [81] proposed a scheduling algo-rithm and took scientific workflows as a model to make fulluse of cloud resources and save energy Shi et al [82]designed a flexible resource allocation and job schedulingmechanism to implement scientific workflows Because thismechanism can implement scientific workflows withinprescribed budgets and deadlines so it can work better thanother mechanisms

Table 5 shows a summary of the abovementioned workse problems in Table 5 are derived from the imple-mentation of the scientific workflow VM deployment refersto the rational allocation of VMs Workflow scheduling refersto reducing the scheduling energy and time In addition itrefers to the scheduling of the workflow on heterogeneoussystems Cost reduction refers to reducing the cost of workflowexecution Effective implementation is about scientific work-flow execution within a specified budget and time

24 Server Optimization Server optimization is also a goodway to save energy e goal of optimizing the server can beachieved by uninstalling unnecessary servers or consoli-dating servers as well as by reasonably scheduling tasksUnlike QoS optimization server optimization aims to op-timize the number of used servers improves the energyefficiency of servers and consolidates servers However QoSoptimization studies how tomake users get better experienceand meet their needs

10 Complexity

241 Cloud Computing Ge et al [83] proposed a game-theoretic method and transformed the problem of mini-mizing energy into a congestion game All mobile devices inthis method are participants in the game e methodchooses a server to unload the computation tasks to optimizeQoS levels and save energy which can optimize the systemand save energy Wang et al [84] proposed a MapReduce-based multitask scheduling algorithm to achieve the ob-jective of energy saving is model is a two-layer modelwhich considers the impact of server performance changeson energy consumption and the limitation of networkbandwidth In addition a local search operator is designedbased on which a two-layer genetic algorithm is proposede algorithm can schedule tens of thousands of tasks incloud and achieve large-scale optimization Yanggratoke

et al [85] proposed a general generic gossip protocol aimingat allocating resources in cloud environment An instanti-ation of this protocol was developed to enable server con-solidation to allocate resources to more servers to meetchanging load patterns

25 LoadBalancing Load balancing can help save energy bymanaging the number of servers and allocating resources

251 Cloud Computing Paya and Marinescu [86] intro-duced an operation model that balances cloud computingload and expands applications to save energy e principleof this model is to define an operating system e systemshould make servers run in the system as many as possible

Table 5 Work summary of scientific workflow execution in cloud computing

Problems Solutions Literatures Advantages

VM deployment A resource allocation method namedEnRealan [76] Performs scientific workflows based on energy perception

across cloud platforms

Workflowscheduling

A scheduling method based on energyperception [77]

Achieves a high parallelism without huge energyconsumption and minimizes the total consumption of

energy and execution time of workflows

A workflow scheduling method with severalobjects and hybrid particle swarm

optimization algorithm[78]

Makes the processors work at any voltage level minimizesthe energy consumption in the process of workflowscheduling and studies the scheduling problem of

workflows on heterogeneous systemsA scheduling algorithm based on various

applications [79] Enables service providers to gain high profits and reducesuser overhead at the same time

Cost reduction A scheduling algorithm based on energyperception [80 81] Minimizes the cost of performing workflows while

meeting the time constraintEffectiveimplementation

A flexible resource allocation and jobscheduling mechanism [82] Implements scientific workflows within prescribed

budgets and deadlines

Table 4 Work summary of resource allocation and management in IoT

Problems Solutions Literatures Advantages

Resourcemanagement

A distributed cloud network framework [66]Replaces the centralized architecture with a distributed cloud

architecture solves the defects of the centralized cloudarchitecture and brings people better experience

A MILP model [68] Assigns the requirement of servicesrsquo resource to heterogeneousnetwork equipment interface

A framework for communication used in5G [69]

Transforms the resource allocation problem into a power andchannel allocation problem minimizes the total energy

consumption and improves QoS levelsA low complexity channel allocation

algorithm [70] Improves throughput of the network and allocates finiteresources to a large group of users

A resource allocation framework withseveral bands under cognitive 5G IoT [71] Manages resources more flexibly and reduces energy

consumption than common single-band approach

A protocol which is distributed andoptimal to allocate resources [72]

Has excellent adaptability when changing topology of networkand dynamically manages resources in the heterogeneous IoT

environmentA multilevel allocating resourcesalgorithm for IoT communication [73] Has fast data processing rate and very low latency in both

saturated and nonsaturated environment

A new algorithm to allocate bandwidthdynamically [74]

Reduces the error of signal reconstruction under the samebandwidth and makes the bandwidth allocation of IoT more

reasonable

A reinforcement learning mechanism [75]Effectively solves the resource allocation problems caused by themismatch of service quality and complex service providing

condition in the IoT

Complexity 11

When no tasks are being performed the system should adjustservers to sleep thus energy consumption can be reducedJustafort et al [87] mainly studied the problem of workloaddistribution across cloud computing environment and pro-posed a method to solve the problem of the VM layout So thefootprint of carbon can be effectively reduced Panwar andMallick [88] proposed an algorithm to dynamically manage theload and effectively distribute the total incoming requestsbetween VMs rough efficient and uniform utilization ofresources this algorithm can achieve uniform distribution ofload between servers Yang et al [89] proposed a powermanagement mechanism to balance the load e system canmonitor VMs and dynamically allocate the resources Yanget al [90] proposed an optimization system to better allocateresources dynamically which can balance the load of VMsrunning on multiple physical machines Under this systemVMs can be migrated automatically to adjust high and lowloads without interrupting services Yang et al [89 90] manageVMs to achieve load balancing ey allocate resources dy-namically to migrate VMs which can balance workloads ondifferent physical machines e difference is that Yanget al[89] integrate a dynamic resource allocation approachwithOpenNebulaWhile Yang et al [90] focuse on avoiding serviceoutages during VM migration

Table 6 shows a summary of the abovementioned workse problems in Table 6 are from the load balancing problemServer management is about the control of the number ofservers running in the system Workload management is therational allocation of workload or tasks VMmanagement refersto configuringVMresources andmigratingVMs to adjust loads

252 Fog Computing Xu et al [91] proposed a methodcalled ldquoDRAMrdquo to dynamically allocate resources in fogcomputing environment which can avoid both too highand too low loads e method first analyzes the loadbalance of different kinds of computing nodes And then itdesigns a fog resource allocation method to achieve loadbalance which allocates resources statically and migratesservices dynamically Oueis et al [92] studied the loadbalance problem in fog computing A custom fog clus-tering algorithm is proposed to solve the problem In theproblem several users need to offload computations andall of their demands need to be handled by local com-puting cluster

253 IoT Wang et al [93] established architecture of theenergy saving targeted system in industrial IoT In additionin order to predict sleep intervals they developed a sleepscheduling and a wake protocol which provide a better wayfor energy saving

3 Conclusion

is paper did a comprehensive study of QoS optimi-zation and energy saving in cloud computing edgecomputing fog computing and IoT models We sum-marized five main problems and analyzed their solutionsproposed by existing works By conducting this surveywe aim to help readers have a deeper understanding onthe concepts of different computing models and thetechniques of QoS optimization and energy saving inthese models

e investigated papers focus on issues about ensuringQoS and reducing SLA violations and resource manage-ment In the case of QoS assurance and SLA violationsreduction the main solution of QoS assurance is efficientVM management is solution can meet customersrsquo re-quirements through reasonable scheduling and integrationof VMs Most of resource management techniques are re-alized by reasonable scheduling of resources which canreduce the waste of VMs servers and traffic

Disclosure

is manuscript is an extension of A Survey of QoS Opti-mization and Energy Saving in Cloud Edge and IoT in e9th EAI International Conference on Cloud Computing

Conflicts of Interest

e authors declare that there are no conflicts of interestregarding the publication of this paper

Acknowledgments

is work was supported by the National Natural ScienceFoundation of China (Grant no 61702274) Natural ScienceFoundation of Jiangsu Province (Grant no BK20170958)and PAPD

Table 6 Work summary of load balancing in cloud computing

Problems Solutions Literatures AdvantagesServermanagement

An operation model that can balance cloudcomputing load and expand application [86] Saves energy by managing the number of servers running in

the system

Workloadmanagement

A hybrid approach [87] Reduces the footprint of carbon and allocates workload acrosscloud computing

An algorithm to dynamically manage theload [88] Manages the load evens the load distribution between

servers and allocates tasks between VMs

VMmanagement

A green power administration mechanism [89] Monitors and jointly allocates VM resources

An optimization system [90]Migrates VMs to adjust high and low loads without

interrupting services and balances the load of VMs runningon multiple physical machines

12 Complexity

References

[1] Q Zhang L Cheng and R Boutaba ldquoCloud computingstate-of-the-art and research challengesrdquo Journal of InternetServices and Applications vol 1 no 1 pp 7ndash18 2010

[2] W Shi J Cao Q Zhang Y Li and L Xu ldquoEdge computingvision and challengesrdquo IEEE Internet of Aings Journal vol 3no 5 pp 637ndash646 2016

[3] M Iorga L Feldman R Barton M J Martin N S Gorenand C Mahmoudi ldquoFog computing conceptual modelrdquo TechRep Recommendations of the National Institute of Standardsand Technology Gaithersburg MD USA 2018

[4] Nebbiolo ldquoFog vs edge computingrdquo Tech Rep NebbioloTechnologies Inc Milpitas CA USA 2018

[5] C T Do N H Tran C Pham M G R Alam J H Son andC S Hong ldquoA proximal algorithm for joint resource allo-cation andminimizing carbon footprint in geo-distributed fogcomputingrdquo in Proceedings of the 2015 International Con-ference on Information Networking (ICOIN) pp 324ndash329IEEE Siem Reap Cambodia January 2015

[6] S Vashi J Ram J Modi S Verma and C Prakash ldquoInternetof things (IoT) a vision architectural elements and securityissuesrdquo in Proceedings of the 2017 International Conference onI-SMAC (IoT in Social Mobile Analytics and Cloud)(I-SMAC) pp 492ndash496 IEEE Coimbatore India February2017

[7] M Mazzucco D Dyachuk and R Deters ldquoMaximizing cloudprovidersrsquo revenues via energy aware allocation policiesrdquo inProceedings of the 2010 IEEE 3rd International Conference onCloud Computing pp 131ndash138 IEEE Miami FL USA July2010

[8] Q He R Zhou X Zhang et al ldquoKeyword search for buildingservice-based systemsrdquo IEEE Transactions on Software En-gineering vol 43 no 7 pp 658ndash674 2016

[9] L Sun H Dong O K Hussain F K Hussain and A X LiuldquoA framework of cloud service selection with criteria inter-actionsrdquo Future Generation Computer Systems vol 94pp 749ndash764 2019

[10] M Mazzucco and D Dyachuk ldquoOptimizing cloud providersrevenues via energy efficient server allocationrdquo SustainableComputing Informatics and Systems vol 2 no 1 pp 1ndash122012

[11] Q He J Han F Chen et al ldquoQos-aware service selection forcustomisable multi-tenant service-based systems maturityand approachesrdquo in Proceedings of the 2015 IEEE 8th Inter-national Conference on Cloud Computing pp 237ndash244 IEEENew York NY USA July 2015

[12] L Sun J Ma H Wang Y Zhang and J Yong ldquoCloud servicedescription model an extension of usdl for cloud servicesrdquoIEEE Transactions on Services Computing vol 11 no 2pp 354ndash368 2015

[13] Y Wang Q He D Ye and Y Yang ldquoFormulating criticality-based cost-effective fault tolerance strategies for multi-tenantservice-based systemsrdquo IEEE Transactions on Software En-gineering vol 44 no 3 pp 291ndash307 2017

[14] S Mustafa K Bilal S U R Malik and S A Madani ldquoSla-aware energy efficient resource management for cloud en-vironmentsrdquo IEEE Access vol 6 pp 15004ndash15020 2018

[15] J Bi H Yuan M Tie and W Tan ldquoSla-based optimisation ofvirtualised resource for multi-tier web applications in clouddata centresrdquo Enterprise Information Systems vol 9 no 7pp 743ndash767 2015

[16] S Singh I Chana and R Buyya ldquoStar sla-aware autonomicmanagement of cloud resourcesrdquo IEEE Transactions on CloudComputing p 1 2017

[17] A Beloglazov and R Buyya ldquoEnergy efficient resourcemanagement in virtualized cloud data centersrdquo in Proceedingsof the 2010 10th IEEEACM International Conference onCluster Cloud and Grid Computing pp 826ndash831 IEEEComputer Society Melbourne Australia May 2010

[18] M Guazzone C Anglano and M Canonico ldquoEnergy-effi-cient resource management for cloud computing infra-structuresrdquo in Proceedings of the 2011 IEEE AirdInternational Conference on Cloud Computing Technology andScience pp 424ndash431 IEEE Athens Greece November 2011

[19] Y Sun J White and S Eade ldquoA model-based system toautomate cloud resource allocation and optimizationrdquo inProceedings of the International Conference on Model DrivenEngineering Languages and Systems pp 18ndash34 SpringerValencia Spain October 2014

[20] G Siddesh and K Srinivasa ldquoSla-driven dynamic resourceallocation on cloudsrdquo in Proceedings of the InternationalConference on Advanced Computing Networking and Securitypp 9ndash18 Springer Surathkal India December 2011

[21] S K Garg S K Gopalaiyengar and R Buyya ldquoSla-basedresource provisioning for heterogeneous workloads in avirtualized cloud datacenterrdquo in Proceedings of the Interna-tional Conference on Algorithms and Architectures for ParallelProcessing pp 371ndash384 Springer Melbourne AustraliaOctober 2011

[22] J Bi Z Zhu and H Yuan ldquoSla-aware dynamic resourceprovisioning for profit maximization in shared cloud datacentersrdquo in Proceedings of the International Conference onHigh Performance Networking Computing and Communi-cation Systems pp 366ndash372 Springer Singapore May 2011

[23] L Qi Y Chen Y Yuan S Fu X Zhang and X Xu ldquoA Qos-aware virtual machine scheduling method for energy con-servation in cloud-based cyber-physical systemsrdquo WorldWide Web vol 4 no 3 pp 1ndash23 2019

[24] A Beloglazov J Abawajy and R Buyya ldquoEnergy-aware re-source allocation heuristics for efficient management of datacenters for cloud computingrdquo Future Generation ComputerSystems vol 28 no 5 pp 755ndash768 2012

[25] L Qi R Wang C Hu S Li Q He and X Xu ldquoTime-awaredistributed service recommendation with privacy-preserva-tionrdquo Information Sciences vol 480 pp 354ndash364 2019

[26] Z Zhai B Cheng Y Tian J Chen L Zhao and M Niu ldquoAdata-driven service creation approach for end-usersrdquo IEEEAccess vol 4 pp 9923ndash9940 2016

[27] L Gu D Zeng S Guo A Barnawi and Y Xiang ldquoCostefficient resource management in fog computing supportedmedical cyber-physical systemrdquo IEEE Transactions onEmerging Topics in Computing vol 5 no 1 pp 108ndash119 2015

[28] L Ni J Zhang C Jiang C Yan and K Yu ldquoResource al-location strategy in fog computing based on priced timed petrinetsrdquo Ieee Internet of Aings Journal vol 4 no 5pp 1216ndash1228 2017

[29] L Wei T Yang F C Delicato et al ldquoOn enabling sustainableedge computing with renewable energy resourcesrdquo IEEECommunications Magazine vol 56 no 5 pp 94ndash101 2018

[30] P Lai Q He G Cui et al ldquoEdge user allocation with dynamicquality of servicerdquo in Proceedings of the International Con-ference on Service-Oriented Computing pp 86ndash101 SpringerToulouse France October 2019

[31] X Xu Q Liu X Zhang J Zhang L Qi and W Dou ldquoAblockchain-powered crowdsourcing method with privacy

Complexity 13

preservation in mobile environmentrdquo IEEE Transactions onComputational Social Systems vol 6 no 6 pp 1407ndash1419 2019

[32] O Rolik E Zharikov and S Telenyk ldquoMicrocloud-basedarchitecture of management system for IoT infrastructuresrdquoin Proceedings of the 2016 Aird International Scientific-Practical Conference Problems of Infocommunications Scienceand Technology (PIC SampT) pp 149ndash151 IEEE KharkivUkaraine October 2016

[33] W He S Guo Y Liang R Ma X Qiu and L Shi ldquoQos-awareand resource-efficient dynamic slicingmechanism for internetof thingsrdquo Computers Materials amp Continua vol 61 no 3pp 1345ndash1364 2019

[34] J Yao and N Ansari ldquoQos-aware fog resource provisioningand mobile device power control in IoT networksrdquo IEEETransactions on Network and Service Management vol 16no 1 pp 167ndash175 2018

[35] W Wang Y Jiang and W Wu ldquoMultiagent-based resourceallocation for energy minimization in cloud computing sys-temsrdquo IEEE Transactions on Systems Man and CyberneticsSystems vol 47 no 2 pp 1ndash16 2016

[36] K Krishnajyothi ldquoParallel data processing for effective dy-namic resource allocation in the cloudrdquo International Journalof Computer Applications vol 70 no 22 pp 1ndash4 2013

[37] M M Hassan B Song M S Hossain and A Alamri ldquoEf-ficient resource scheduling for big data processing in cloudplatformrdquo in Proceedings of the International Conference onInternet and Distributed Computing Systems pp 51ndash63Springer Calabria Italy September 2014

[38] C-M Wu R-S Chang and H-Y Chan ldquoA green energy-efficient scheduling algorithm using the dvfs technique forcloud datacentersrdquo Future Generation Computer Systemsvol 37 no 7 pp 141ndash147 2014

[39] Z Wang and X Su ldquoDynamically hierarchical resource-al-location algorithm in cloud computing environmentrdquo AeJournal of Supercomputing vol 71 no 7 pp 2748ndash2766 2015

[40] W-Y Lin G-Y Lin and H-Y Wei ldquoDynamic auctionmechanism for cloud resource allocationrdquo in Proceedings ofthe 2010 10th IEEEACM International Conference on ClusterCloud and Grid Computing pp 591-592 IEEE ComputerSociety Victoria Australia May 2010

[41] Y O Yazir C Matthews R Farahbod et al ldquoDynamic re-source allocation in computing clouds using distributedmultiple criteria decision analysisrdquo in Proceedings of the 2010IEEE 3rd International Conference on Cloud Computingpp 91ndash98 IEEE Washington DC USA July 2010

[42] W Lin J Z Wang C Liang and D Qi ldquoA threshold-baseddynamic resource allocation scheme for cloud computingrdquoProcedia Engineering vol 23 no 5 pp 695ndash703 2011

[43] X Xu S Fu L Qi et al ldquoAn IoT-oriented data placementmethod with privacy preservation in cloud environmentrdquoJournal of Network and Computer Applications vol 124pp 148ndash157 2018

[44] J Bokyun P Md Jalil L Daeho and S Doug Young ldquoEf-ficient computation offloading in mobile cloud computing forvideo streaming over 5 grdquo Computers Materials amp Continuavol 61 no 2 pp 439ndash463 2019

[45] H Sarbazi-Azad and A Y Zomaya ldquoEnergy-efficient resourceutilization in cloud computingrdquo in Large Scale Network-Centric Distributed Systems pp 377ndash408 Wiley-IEEE PressHoboken NJ USA 2014

[46] C-H Hsu K D Slagter S-C Chen and Y-C ChungldquoOptimizing energy consumption with task consolidation incloudsrdquo Information Sciences vol 258 no 3 pp 452ndash4622014

[47] C-H Hsu S-C Chen C-C Lee et al ldquoEnergy-aware taskconsolidation technique for cloud computingrdquo in Proceedingsof the 2011 IEEE Aird International Conference on CloudComputing Technology and Science pp 115ndash121 IEEE Ath-ens Greece 2011

[48] S K Panda and P K Jana ldquoAn efficient task consolidationalgorithm for cloud computing systemsrdquo in Proceedings of theInternational Conference on Distributed Computing and In-ternet Technology pp 61ndash74 Springer Bhubaneswar IndiaJanuary 2016

[49] L Yin J Luo and H Luo ldquoTasks scheduling and resourceallocation in fog computing based on containers for smartmanufacturingrdquo IEEE Transactions on Industrial Informaticsvol 14 no 10 pp 4712ndash4721 2018

[50] M Aazam and E-N Huh ldquoDynamic resource provisioningthrough fog micro datacenterrdquo in Proceedings of the 2015IEEE International Conference on Pervasive Computing andCommunication Workshops (PerCom Workshops) pp 105ndash110 IEEE St Louis MO USA March 2015

[51] B Jia H Hu Y Zeng T Xu and Y Yang ldquoDouble-matchingresource allocation strategy in fog computing networks basedon cost efficiencyrdquo Journal of Communications and Networksvol 20 no 3 pp 237ndash246 2018

[52] H Zhang Y Xiao S Bu D Niyato F R Yu and Z HanldquoComputing resource allocation in three-tier IoT fog net-works a joint optimization approach combining stackelberggame and matchingrdquo IEEE Internet of Aings Journal vol 4no 5 pp 1204ndash1215 2017

[53] J Tan T-H Chang and T Q Quelc ldquoMinimum energyresource allocation in fog radio access network with fronthauland latency constraintsrdquo in Proceedings of the 2018 IEEE 19thInternational Workshop on Signal Processing Advances inWireless Communications (SPAWC) pp 1ndash5 IEEE KalamataGreece June 2018

[54] D R de Vasconcelos R M de Castro Andrade andJ N de Souza ldquoSmart shadowndashan autonomous availabilitycomputation resource allocation platform for internet ofthings in the fog computing environmentrdquo in Proceedings of the2015 International Conference on Distributed Computing inSensor Systems pp 216-217 IEEE Fortaleza Brazil June 2015

[55] M Aazam M St-Hilaire C-H Lung and I LambadarisldquoPre-fog IoT trace based probabilistic resource estimation atfogrdquo in Proceedings of the 2016 13th IEEE Annual ConsumerCommunications amp Networking Conference (CCNC) pp 12ndash17 IEEE Las Vegas NV USA January 2016

[56] N D Tung L L Bao and B Vijay ldquoPrice-based resourceallocation for edge computing a market equilibrium ap-proachrdquo IEEE Transactions on Cloud Computing p 1 2018

[57] X Xu C He Z Xu L Qi S Wan and M Z A BhuiyanldquoJoint optimization of offloading utility and privacy for edgecomputing enabled IoTrdquo IEEE Internet of Aings Journal2019

[58] X Xu Y Chen X Zhang Q Liu X Liu and L Qi ldquoAblockchain-based computation offloading method for edgecomputing in 5G networksrdquo Software Practice and Experi-ence 2019

[59] X Xu Y Li T Huang et al ldquoAn energy-aware computationoffloading method for smart edge computing in wirelessmetropolitan area networksrdquo Journal of Network and Com-puter Applications vol 133 pp 75ndash85 2019

[60] X Xu Y Xue L Qi et al ldquoAn edge computing-enabledcomputation offloading method with privacy preservation forinternet of connected vehiclesrdquo Future Generation ComputerSystems vol 96 pp 89ndash100 2019

14 Complexity

[61] G Yeting L Fang X Nong and C Zhengguo ldquoTask-basedresource allocation bid in edge computing micro datacenterrdquoComputers Materials amp Continua vol 61 no 2 pp 777ndash7922019

[62] X Chen L Jiao W Li and X Fu ldquoEfficient multi-usercomputation offloading for mobile-edge cloud computingrdquoIEEEACM Transactions on Networking vol 24 no 5pp 2795ndash2808 2016

[63] B Gao L He X Lu C Chang K Li and K Li ldquoDevelopingenergy-aware task allocation schemes in cloud-assisted mo-bile workflowsrdquo in Proceedings of the 2015 IEEE InternationalConference on Computer and Information Technology Ubiq-uitous Computing and Communications Dependable Auto-nomic and Secure Computing Pervasive Intelligence andComputing pp 1266ndash1273 IEEE Liverpool UK October2015

[64] X Xu X Zhang H Gao Y Xue L Qi andW Dou ldquoBecomeblockchain-enabled computation offloading for IoT in mobileedge computingrdquo IEEE Transactions on Industrial Infor-matics 2019

[65] W Yifei W Zhaoying G Da and Y F Richard ldquoDeepq-learning based computation offloading strategy for mobileedge computingrdquo Computers Materials amp Continua vol 59no 1 pp 89ndash104 2019

[66] M Barcelo A Correa J Llorca A M Tulino J L Vicarioand A Morell ldquoIoT-cloud service optimization in nextgeneration smart environmentsrdquo IEEE Journal on SelectedAreas in Communications vol 34 no 12 pp 4077ndash40902016

[67] B Cheng M Wang S Zhao Z Zhai D Zhu and J ChenldquoSituation-aware dynamic service coordination in an IoTenvironmentrdquo IEEEACM Transactions On Networkingvol 25 no 4 pp 2082ndash2095 2017

[68] V Angelakis I Avgouleas N Pappas and D Yuan ldquoFlexibleallocation of heterogeneous resources to services on an IoTdevicerdquo in Proceedings of the 2015 IEEE Conference onComputer Communications Workshops (INFOCOMWKSHPS) pp 99-100 IEEE Hong Kong China April 2015

[69] S Li Q Ni Y Sun G Min and S Al-Rubaye ldquoEnergy-ef-ficient resource allocation for industrial cyber-physical IoTsystems in 5g erardquo IEEE Transactions on Industrial Infor-matics vol 14 no 6 pp 2618ndash2628 2018

[70] X Liu Z Qin Y Gao and J A McCann ldquoResource allocationin wireless powered IoT networksrdquo IEEE Internet of AingsJournal vol 6 no 3 pp 4935ndash4945 2019

[71] W Ejaz and M Ibnkahla ldquoMulti-band spectrum sensing andresource allocation for IoT in cognitive 5g networksrdquo IEEEInternet of Aings Journal vol 5 no 1 pp 150ndash163 2017

[72] G Colistra V Pilloni and L Atzori ldquoTask allocation in groupof nodes in the IoT a consensus approachrdquo in Proceedings ofthe 2014 IEEE International Conference on Communications(ICC) pp 3848ndash3853 IEEE Sydney Australia June 2014

[73] J Li Q Sun and G Fan ldquoResource allocation for multiclassservice in IoTuplink communicationsrdquo inProceedings of the 20163rd International Conference on Systems and Informatics (ICSAI)pp 777ndash781 IEEE Shanghai China November 2016

[74] L I Zheng and K H Liu ldquoDynamic bandwidth resourceallocation algorithm in internet of things and its applicationrdquoComputer Engineering vol 38 no 17 pp 16ndash19 2012

[75] K Gai and M Qiu ldquoOptimal resource allocation using re-inforcement learning for IoT content-centric servicesrdquo Ap-plied Soft Computing vol 70 pp 12ndash21 2018

[76] X Xu W Dou X Zhang and J Chen ldquoEnreal an energy-aware resource allocation method for scientific workflow

executions in cloud environmentrdquo IEEE Transactions onCloud Computing vol 4 no 2 pp 166ndash179 2016

[77] K Bousselmi Z Brahmi and M M Gammoudi ldquoEnergyefficient partitioning and scheduling approach for scientificworkflows in the cloudrdquo in Proceedings of the 2016 IEEEInternational Conference on Services Computing (SCC)pp 146ndash154 IEEE San Francisco CA USA June 2016

[78] Y Sonia C Rachid K Hubert and G Bertrand ldquoMulti-objective approach for energy-aware workflow scheduling incloud computing environmentsrdquo Ae Scientific World Jour-nal vol 2013 Article ID 350934 13 pages 2013

[79] F Cao Efficient Scientific Workflow Scheduling in CloudEnvironment Southern Illinois University Carbondale ILUSA 2014

[80] Z Li J Ge H Hu W Song H Hu and B Luo ldquoCost andenergy aware scheduling algorithm for scientific workflowswith deadline constraint in cloudsrdquo IEEE Transactions onServices Computing vol 11 no 4 pp 713ndash726 2015

[81] M Khaleel and M M Zhu ldquoEnergy-aware job managementapproaches for workflow in cloudrdquo in Proceedings of the 2015IEEE International Conference on Cluster Computingpp 506-507 IEEE Chicago IL USA September 2015

[82] J Shi J Luo F Dong and J Zhang ldquoA budget and deadlineaware scientific workflow resource provisioning and sched-uling mechanism for cloudrdquo in Proceedings of the 2014 IEEE18th International Conference on Computer Supported Co-operative Work in Design (CSCWD) pp 672ndash677 IEEEHsinchu Taiwan January 2014

[83] Y Ge Y Zhang Q Qiu and Y-H Lu ldquoA game theoreticresource allocation for overall energy minimization in mobilecloud computing systemrdquo in Proceedings of the 2012ACMIEEE International Symposium on Low Power Electronicsand Design pp 279ndash284 ACM Redondo Beach CA USAAugust 2012

[84] X Wang Y Wang and C Yue ldquoAn energy-aware bi-leveloptimization model for multi-job scheduling problems undercloud computingrdquo Soft Computing vol 20 no 1 pp 303ndash3172016

[85] R Yanggratoke F Wuhib and R Stadler ldquoGossip-basedresource allocation for green computing in large cloudsrdquo inProceedings of the 2011 7th International Conference onNetwork and Service Management pp 1ndash9 IEEE ParisFrance October 2011

[86] A Paya and D C Marinescu ldquoEnergy-aware load balancingand application scaling for the cloud ecosystemrdquo IEEETransactions on Cloud Computing vol 5 no 1 pp 15ndash272015

[87] V D Justafort R Beaubrun and S Pierre ldquoA hybrid ap-proach for optimizing carbon footprint in intercloud envi-ronmentrdquo IEEE Transactions on Services Computing vol 12no 2 pp 186ndash198 2016

[88] R Panwar and B Mallick ldquoLoad balancing in cloud com-puting using dynamic load management algorithmrdquo inProceedings of the 2015 International Conference on GreenComputing and Internet of Aings (ICGCIoT) pp 773ndash778IEEE Noida India October 2015

[89] C-T Yang K-C Wang H-Y Cheng C-T Kuo andW C C Chu ldquoGreen power management with dynamicresource allocation for cloud virtual machinesrdquo in Proceedingsof the 2011 IEEE International Conference on High Perfor-mance Computing and Communications pp 726ndash733 IEEEAlberta Canada September 2011

[90] C-T Yang H-Y Cheng and K-L Huang ldquoA dynamicresource allocation model for virtual machine management

Complexity 15

on cloudrdquo in Proceedings of the International Conference onGrid and Distributed Computing pp 581ndash590 SpringerGangneug Korea December 2011

[91] X Xu F Shucun C Qing et al ldquoDynamic resource allocationfor load balancing in fog environmentrdquo Wireless Commu-nications amp Mobile Computing vol 2018 no 2 Article ID6421607 15 pages 2018

[92] J Oueis E C Strinati and S Barbarossa ldquoe fog balancingload distribution for small cell cloud computingrdquo in Pro-ceedings of the 2015 IEEE 81st Vehicular Technology Confer-ence (VTC Spring) pp 1ndash6 IEEE Glasgow UK May 2015

[93] K Wang Y Wang Y Sun S Guo and J Wu ldquoGreen in-dustrial internet of things architecture an energy-efficientperspectiverdquo IEEE Communications Magazine vol 54 no 12pp 48ndash54 2016

16 Complexity

life of consumers He et al [33] proposed a dynamic networkslice strategy e network slice can be dynamically adjustedaccording to the time-varying resource demands is methodcan improve the utilization of the underlying resources andbetter meet different QoS demands Yao and Ansari [34]proposed an algorithm to determine the number of VMs to berented and to control the power supply us the cost of thesystem can be minimized and the QoS can be improvedFormula (5) [34] is used to limit the delay requirement of QoSe total delay must not exceed the computation deadline ofeach task and the total delay is composed of wireless trans-mission delay and fog processing delay

tci + t

wi leDi foralliϵN (5)

where c and w respectively represents fog processing andwireless transmission i denotes a location tc

i represents thedelay of processing tw

i represents the delay of wirelesstransmission Di denotes the deadline and N denotes dif-ferent locations

22 Resource Management and Allocation Rational alloca-tion of resources is an effective means to save energy

221 Cloud Computing Wang et al [35] introduced anallocation method based on distributed multiagent to al-locate VMs to physical machines e method can realize

VM consolidation and consider the migration costs si-multaneously In addition a VM migration mechanismbased on local negotiation is proposed to avoid unnecessaryVM migration costs Hassan et al [37] established a for-mulation of universal problem and proposed a heuristicalgorithm which has optimal parameters Under this for-mulation dynamic resource allocation can be made to meetthe QoS requirements of applications And the cost neededfor dynamic resource allocation can be minimized with thisalgorithm Wu et al [38] proposed a scheduling algorithmbased on the technology that can scale the voltage frequencydynamically in cloud computing e algorithm can allocateresources for performing tasks and realize low powerconsumption network infrastructure Compared with otherschemes this scheme not only sacrifices the performance ofexecution operations but also saves more energy

Sarbazi and Zomaya [45] used two job consolidationheuristic methods to save energy One is MaxUtil to betterutilize resources and the other is Energy-Conscious TaskConsolidation to reduce energy consumption ese twomethods can promote the concurrent execution of multipletasks and improve the energy efficiency Hsu et al [46]proposed a job consolidation technique to minimize energyconsumption Formula (6) [46] defines the energy con-sumption of VM Vi from time t0 to tm in the cluster isdefined And formula (7) [46] defines the total energyconsumption in a virtual cluster VCk in the period Inaddition the proposed technique limits the CPU usage and

Reference machine

Reference machine

SLA

SLA

Application A1

Application An

TIER 1

TIER 2

TIER 3

TIER 1

TIER 2

TIER 3

Cloud infrastructure

VM

VM

VM

VM

VM

VM

Performance measuresEnergy consumptions

Resource costs

Resourcemanager

Migrationmanager

Physicalmachinemanagers

Applicationmanagers

Figure 4 Framework of the proposed three-fold system [18]

Complexity 5

Tabl

e1

Worksummaryof

QoS

guaranteeing

orSL

Aassurancein

clou

dcompu

ting

Subp

roblem

sSolutio

nsLiteratures

Advantages

Server

managem

ent

Apo

licyof

dynamicallocatio

nof

poweringserverss

witches

[71

0]Maxim

izes

benefitsim

proves

QoSa

ndminim

izes

power

consum

ption

Workloads

consolidation

Atechniqu

eforconsolidatingworkloads

[14]

Achievesenergy

saving

sby

usingfewestserverswhile

redu

cing

SLA

violations

VM

managem

ent

Anarchitecturethat

canadministrate

itselfa

ndamixed

queuingmod

el[15]

Makes

morerevenu

esandmeets

different

requ

irem

ents

ofcustom

ersanddecidesthe

numberof

VMsforeach

layerof

avirtuala

pplication

Asystem

tointegrateanddynamically

redistribu

teVMs

[17]

Achievesthe

goalofsaving

energy

throughVM

integrationandprovidesah

ighQoS

levelat

thesametim

e

AQoS-awareVM

schedu

lingstrategy

named

QVMS

[23]

Effectiv

elymanages

resourcesin

thenetworkph

ysical

system

toredu

cetheenergy

consum

ptionandim

proves

QoS

AVM

integrationmetho

dwith

severaltargets

[24]

Savesenergy

andredu

cesSL

Aviolations

byapplying

different

strategies

todifferent

load

states

oftheho

st

Self-managem

ent

Atechno

logy

named

STAR

[16]

Redu

cesSL

Aviolations

andim

proves

paym

ente

fficiency

ofclou

dservices

Adynamic

resource

managem

entsystem

[18]

Self-manages

theresourceso

fcloud

infrastructurestoprovidea

ppropriateQoS

andfitsthe

changing

workloads

dynamically

Resource

managem

ent

Amod

elthat

canrealizetheindepend

entallocatio

nof

resources

[19]

Obtains

theop

timal

resource

confi

guratio

nmeets

theQoS

requ

irem

entsand

provides

econ

omical

clou

dresources

Adynamic

resource

allocatio

nstrategy

[20

21]

Redu

cesSL

Aandmaxim

izes

revenu

esandresource

utilizatio

non

theclou

d

Amixed

queuemod

el[22]

Reason

ably

confi

gurestheresourcesin

theclou

ddata

centerimproves

thesystem

performancereduces

theadditio

nalcosto

fusin

gresourcesmeetstherequ

ired

QoSand

provides

virtualresou

rces

toeach

layerof

virtuala

pplicationservices

Service

managem

ent

Aun

ified

semantic

mod

elthat

candescribe

clou

dservice

[12]

Improves

theability

ofmod

elon

servicerank

ingandenriches

thelang

uage

expressio

n

Arecommendatio

nservicestrategy

[25]

Emph

asizes

theinflu

ence

oftim

efactorson

QoS

andim

proves

thetradition

allocatio

n-sensitive

hash

techno

logy

toprotectusersrsquoprivacy

Aclou

dserviceselectionmetho

dusingfuzzymeasure

and

Cho

quet

integral

andafram

eworkbasedon

priority

[9]

Selectsservicewhenhistorical

inform

ationisinsufficientto

determ

inethecriteria

relatio

nships

andweigh

ts

reeserviceselectionmetho

dsthat

supp

ortQ

oSandcan

combine

multitenants

ervice-based

system

s[11]

Achievesthree

degreeso

fmultitenantm

aturity

which

ismoreeffi

cientthanthetradition

alsin

gle-user

approach

Afault-tolerant

strategy

basedon

multitenants

ervice

criticality

[13]

Guaranteesthequ

ality

ofthemultitenant-basedservicesystem

Aservice-basedsystem

supp

ortin

gkeyw

ordsearch

[8]

Effectiv

elyhelpssystem

engineerswho

areno

tfam

iliar

with

service-oriented

architecture

techno

logy

tobu

ildservice-oriented

system

s

6 Complexity

merges tasks in virtual clusters Once a task migrationhappens the energy cost model will take into account thenetwork latency Sarbazi-Azad and Zomaya [45] and Hsuet al [46] both maximize the benefit of cloud resources byusing task merging techniques Sarbazi-Azad and Zomaya[45] uses a greedy algorithm called MaxUtil While Hsuet al [46] takes into account the network latency associatedwith task migration So in [46] a 17 improvement isachieved over MaxUtil

E0m Vi( 1113857 1113944m

t0Et Vi( 1113857 (6)

E0m VCk( 1113857 1113944n

i0E0m Vi( 1113857 (7)

where Et is the energy consumption in unit time and n is thenumber of VMs in the cluster

Hsu et al [47] proposed a task integration technologybased on the energy perception According to the charac-teristics of most cloud systems the principle of using 70CPU is proposed to administrate job integration amongvirtual clusters is technology is very effective in reducingthe amount of energy consumed in cloud systems bymerging tasks Panda and Jana [48] proposed an algorithmwith several criteria to combine tasks e algorithm notonly considers the time needed for processing jobs but alsoconsiders the utilization rate of VMs And the algorithm ismore energy efficient because it takes into account not onlythe processing time but also the utilization rate of VMsWang and Su [39] proposed a resource allocation algorithmto deal with wide range of communication between nodes incloud environment is algorithm uses recognition tech-nology to dynamically distribute jobs and nodes accordingto computing ability and factors of storage And it can re-duce the traffic when allocating resources because it usesdynamic hierarchy Lin et al [40] proposed a dynamicauction approach for resource allocation e approach canensure that even if there are many users and resourcesproviders will have reasonable profits and computing re-sources will be allocated correctly Yazir et al [41] proposeda new method to manage resources dynamically and au-tonomously Firstly resource management is split into jobsand each job is executed by autonomous nodes Secondautonomous nodes use the method called PROMETHEE toconfigure resources Krishnajyothi [36] proposed a frame-work which can implement parallel task processing to solvethe problem of low efficiency when submitting large tasksCompared with the static framework this framework candynamically allocate VMs thus reducing costs and the timeof processing tasks Lin et al [42] proposed a method toallocate resources dynamically by using thresholds Becausethis method uses the threshold value it can optimize thereallocation of resources improve the usage of resourcesand reduce the cost Xu et al [43] proposed a data placementstrategy named IDP for the data generated by IoTs devices toachieve reasonable data placement In this way the privacyof these data can be protected while resources are allocatedreasonably Jo et al [44] proposed a computing offload

framework under 5G network e framework transfers thecomputing burden to the cloud thus reducing the com-puting load of clients and the communication cost

Table 2 shows a summary of the abovementioned workse problem of resource allocation and management incloud computing can be divided into problems in Table 2VM management is about a reasonable configuration ofVMs Resource allocation represents the dynamic andflexible allocation of resources Task integration refers tocombining tasks to save energy and improve efficiency

222 Fog Computing Yin et al [49] established a newmodel of scheduling jobs which applies containers In orderto make sure that jobs can be finished on time a jobscheduling algorithm is developed e algorithm can alsooptimize the number of tasks that can be performed togetheron the nodes in fog computing And this paper proposes aredistribution mechanism to shorten the delay of tasksese methods are very effective in reducing task delaysAazam and Huh [50] established a framework to admin-istrate resources effectively in the mode of fog computingConsidering that there are various types of objects anddevices the connection between them may be volatile So amethod for predicting and administrating resources isproposed e method considers that any object or devicecan quit using resources at anytime Cuong et al [5] studiedthe allocating resources jointly problem and the problem ofcarbon footprint minimization in fog data center Formula(8) [5] is used to denote the energy consumption of serversIn addition a distributed algorithm is proposed to solve theproblem of wide range optimization

P(y) C middot Pidle + Ppeak minus Pidle1113872 1113873 middot y middot κ (8)

where P(y) represents the power supply required by theservers in a data center y represents the video stream κdenotes a conversion factor that converts the video streaminto workload C represents the data centerrsquos load capacityand Pidle and Ppeak respectively represent the idle power andpeak power of the servers

Jia et al [51] studied the problem of computing resourceallocation in fog computing network with three levelsFirstly the problem of resource allocation is transformedinto a bilateral matching optimal problem And then a bi-matching approach is proposed for this problem which canimprove the performance of the system and obtain highercost efficiency Zhang et al [52] proposed a framework forjoint optimization under fog computing to allocate fognodesrsquo finite computing resources e framework canachieve the best allocation and effectively improve thenetworksrsquo performance Tan et al [53] presented a method toallocate computing and communication resources emethod transfers computing jobs to remote cloud and nodesand simplifies edge nodesrsquo computing and computing en-ergy Vasconcelos et al [54] developed a platform to allocateresources accessible to client devices in fog computingenvironment allocating the resources of devices near thehost to meet the applications needs for rapid response tocomputing resources Aazam et al [55] presented a method

Complexity 7

to estimate and manage resource in fog computing emethod is based on the fluctuation of customer abandon-ment probability type and price of service and so on

Table 3 shows a summary of the abovementioned workse problems in Table 3 are also derived from resourceallocation and management problems Task allocationrepresents the scheduling and redistribution of tasks Re-source allocation is still about the dynamic and flexibleallocation of resources Low latency refers to taking shorttime to configure and manage resources which can improveefficiency

223 Edge Computing Tung et al [56] proposed a newframework for resource allocation based on market needse resources come from edge nodes (ENs) with limitedheterogeneous capabilities and are allocated to multiplecompeting services on the network edge Generating amarket equilibrium solution by reasonably pricing ENs canobtain the maximum utilization of marginal computingresources Xu et al [57] proposed a strategy to optimizeoffloading and privacy protection is strategy shifts tasksfirstly to improve the resource utilization of resource-limited

edge cells And then it balances QoS performance andprivacy protection to achieve joint optimization

Xu et al [58] proposed an offload strategy for edgecomputing under 5G network which uses block chaintechnologye optimal strategy is further obtained by usingthe balanced offloading method It solves the problem ofdata loss under the condition of transmission delay which iscaused by the uneven requirements of user equipments onresources Xu et al [59] proposed a computational off-loading method named EACO to reduce the energy con-sumption in smart computing models Figure 5 showsarchitecture of smart edge computing where the shortestpath is used to unload tasks EACO uses genetic algorithmsto reduce the energy consumption for operating edgecomputing nodes and improve the efficiency of performingcomplex computing tasks Xu et al [60] proposed a com-putational offloading strategy for edge computing to protectthe privacy of interconnected vehicle networks ey firstlyanalyzed privacy conflicts of tasks And then they designedthe communication route to obtain routing vehicles whichcan achieve the optimization of several objectives Yetinget al [61] proposed a unique resource allocation mechanisme mechanism takes each individual task as the basis for

Table 2 Work summary of resource allocation and management in cloud computing

Subproblems Solutions Literatures Advantages

VMmanagement

A VM allocation method based on severaldistributed agents [35] Centralizes VMs to physical machines and reduces the

overall energy costA framework which can implement parallel

task processing [36] Dynamically allocates VMs for large tasks

Resourceallocation

A general problem formula and a heuristicalgorithm with optimization parameters [37]

Dynamically allocates resources according to QoSrequirements and realizes energy saving by optimizing the

number of serversA scheduling algorithm based on thetechnology that can scale the voltage

frequency dynamically[38] Ensures the performance of executing jobs while

implementing green computing

A fuzzy pattern recognition technology [39] Reduces the traffic when allocating resources and usesdynamic hierarchy

A dynamic auction approach [40]Guarantees profits of providers and allocates resourcescorrectly when there are a large amount of users and

resources

PROMETHEE [41]Allows node agents to decompose and execute tasksautonomously to improve the flexibility of resource

allocation

A method using thresholds [42]Optimizes resource reallocation improves resource usagereduces cost and studies resource allocation strategies at the

application level

IoT-oriented data placement (IDP) method [43] Protects data privacy allocates resources reasonably andfocuses on the placement method of IoT data

A computing offload framework under 5Gnetwork [44] Reduces the computing load of clients and the

communication cost

Taskconsolidation

Two job consolidation heuristics namedldquoMaxUtilrdquo and energy-conscious task

consolidation[45] Promotes concurrent execution of multiple tasks and

improves energy efficiency

A job consolidation technique aiming atenergy saving [46 47]

Reduces the power consumption of cloud system protectsdata privacy allocates resources reasonably limits CPU

usage and merges tasks in virtual clusters

An algorithm with several criteria [48]Takes into account the job processing time and the VM

utilization rate simultaneously Dramatically reduces energyconsumption compared with state-of-the-art works

8 Complexity

resource allocation rather than for the whole service It reducesthe packet loss rate and saves energy by unloading services

224 MEC Chen et al [62] studied the problem of com-puting unloading with several users in the environment ofMEC with wireless interference which have many channels Inaddition a distributed algorithm for computing unloading isdevelopede algorithm can perform the unloading well evenwhen there are a large number of users Gao et al [63] built aquadratic binary program which is able to assign tasks inmobile cloud computing environment Two algorithms arepresented to obtain the optimal solution Both of these heuristicalgorithms can effectively solve the task assignment problemXu et al [64] proposed an offloadingmethod using block chaintechnology It can guarantee the loss of data in offloading tasksunder edge computing And it can solve the problem of re-source requests out of proportion due to the limited load of

edge computing equipment during task transfer Yifei et al [65]proposed a model-free reinforcement learning framework tosolve the problem of computational unloadingis model canbe applied to the computational unloading with time-changingcomputing requests

225 IoT Barcelo et al [66] expressed the problem ofservice allocation [67] as a mixed flow problem with min-imum cost which can be solved by LP solving this serviceallocation problem can solve the problems of unbalancednetwork load and delay of end-to-end service And it canalso figure out the problem of excessive consumption ofelectricity brought by the architecture of centralized cloudAngelakis et al [68] assigned the requirements of servicesresources to heterogeneous network interfaces of equipmentsSo more heterogeneous network interfaces can be used by alarge amount of services

Mobile devices

Mobile devices ECN

ECN

ECNECN

AP

AP

AP

APAP

Figure 5 An architecture of smart edge computing [59]

Table 3 Work summary of resource allocation and management in fog computing

Problems Solutions Literatures AdvantagesTaskallocation

A new model of scheduling jobs and aredistribution mechanism [49] Finishes jobs on time optimizes the number of tasks and

shortens the delay of tasks

Resourceallocation

A framework to administrate resources and amethod to predict and administrate resources [50]

Assists service providers to predict the amount of availableresources based on different types of service customers anddeals with the phenomenon that objects or devices withdraw

from resource utilization at any time

A bi-matching approach [51] Improves the performance of the system and obtains highercost efficiency

A framework for joint optimization [52] Optimizes resource allocation and improves networkperformance

A method to allocate computing andcommunication resources [53]

Allocates computing and communication resources transferscomputing jobs to remote cloud and nodes simplifies edge

nodesrsquo computing and saves computing energyA platform to allocate resources accessible to

client devices [54] Enables rapid response to computing resources and allocatesthe resources of devices near the host

A framework to manage resources [55] Considers the fluctuation of the customer abandonmentprobability and service types

Low latency A distributed algorithm [5] Solves the problem of wide range optimization and allocatesresources jointly

Complexity 9

Li et al [69] proposed communication framework in 5Gand studied the problem of allocating power andchannels So the signal data in the channel can beavailable and the total energy efficiency can be maximumFormula (9) [69] shows how to calculate the energy ef-ficiency of a system

U 1113944K

k11113944

M

i1EES

iK + 1113944K

K11113944

N

i1EEA

jk (9)

where EESiK and EEA

jk respectively denote the energy ef-ficiency of sensor S and actuator A on channels e sets ofsensors actuators and channels are respectively repre-sented as s S1 S2 SM1113864 1113865 A A1 A2 AN1113864 1113865 andC C1 C2 CK1113864 1113865

Liu et al [70] studied the problem of allocating resourcesefficiently on IoT that supplies wireless power In thismethod users are first grouped into accessible channelsAnd then power distribution of users grouped in the samechannel is studied to improve throughput of the networkis method can allocate finite resources to a large group ofusers Ejaz and Ibnkahla [71] proposed the resource allo-cation framework with several bands under cognitive 5GIoT In the highly dynamic environment of the IoTmultiband method can manage resources more flexiblyand reduce more energy consumption In addition areconstruction approach with several levels is proposed toallocate resources reasonably for applications with dif-ferent needs of QoS Colistra et al [72] proposed aprotocol which is distributed and optimal to allocateresources in heterogeneous IoT Because this protocol hasexcellent adaptability when changing topology of net-work it can distribute resources evenly among nodes Jianet al [73] proposed a multilevel allocating resources al-gorithm for IoT communication using advanced tech-nology e algorithm uses hierarchical structure and hasfast data processing rate and very low latency in a satu-rated or not saturated environment

Zheng and Liu [74] proposed a new algorithm to allocatebandwidth dynamically for controlling remote computers inthe IoT is method can reduce the error of signal re-construction under the same bandwidth and make thebandwidth allocation of IoT more reasonable Gai and Qiu[75] used reinforcement learning mechanisms to allocateresources to achieve high Quality of Experience ismethod can effectively solve the resource allocation prob-lems caused by the mismatch of service quality and complexservice providing condition in the IoT

Table 4 shows a summary of the above works eproblem in Table 4 represents the realization of dynamic andflexible allocation of resources e resources here canrepresent channels bandwidth and power

23 Scientific Workflow Execution Implementing scientificworkflows especially in heterogeneous environments canreduce resource waste and reduce energy costs Scientificworkflow can be obtained by reasonably allocating resourcesand dynamically deploying VMs

231 Cloud Computing Xu et al [76] proposed a resourceallocation method called EnRealan to solve the problem ofenergy consumption e dynamic deployment of VMs isgenerally adopted to execute scientific workflows Bousselmiet al [77] proposed a scheduling method based on energyperception for executing scientific workflows in cloudcomputing At first an algorithm of splitting workflows forenergy minimization is presented which can achieve a highparallelism without huge energy consumption en aheuristic algorithm used to optimize cat swarm is proposedfor the created partitions e algorithm can minimize thetotal consumption of energy and the execution time ofworkflows Sonia et al [78] proposed a workflow schedulingmethod with several objects and hybrid particle swarmoptimization algorithm In addition a method for dy-namically scaling voltage and frequency is proposed emethod canmake the processors work at any voltage level soas to minimize the energy consumption in the process ofworkflow scheduling Both Bousselmi et al [77] and Soniaet al [78] use scheduling method to achieve scientificworkflows and study the problem of energy consumptione difference is that Bousselmi et al [77] focuses on in-tensive computing tasks while Sonia et al [78] focuses onworkflow scheduling on heterogeneous computing systems

Cao [79] established a scheduling algorithm of scientificworkflows with an objective of energy savingis algorithmcan enable service providers to gain high profits and reduceusersrsquo overhead at the same time Li et al [80] proposed ascheduling algorithm based on cloud computing which canminimize cost of performing workflows within a specifiedtime In addition the rented VM was modified to save costfurther Khaleel and Zhu [81] proposed a scheduling algo-rithm and took scientific workflows as a model to make fulluse of cloud resources and save energy Shi et al [82]designed a flexible resource allocation and job schedulingmechanism to implement scientific workflows Because thismechanism can implement scientific workflows withinprescribed budgets and deadlines so it can work better thanother mechanisms

Table 5 shows a summary of the abovementioned workse problems in Table 5 are derived from the imple-mentation of the scientific workflow VM deployment refersto the rational allocation of VMs Workflow scheduling refersto reducing the scheduling energy and time In addition itrefers to the scheduling of the workflow on heterogeneoussystems Cost reduction refers to reducing the cost of workflowexecution Effective implementation is about scientific work-flow execution within a specified budget and time

24 Server Optimization Server optimization is also a goodway to save energy e goal of optimizing the server can beachieved by uninstalling unnecessary servers or consoli-dating servers as well as by reasonably scheduling tasksUnlike QoS optimization server optimization aims to op-timize the number of used servers improves the energyefficiency of servers and consolidates servers However QoSoptimization studies how tomake users get better experienceand meet their needs

10 Complexity

241 Cloud Computing Ge et al [83] proposed a game-theoretic method and transformed the problem of mini-mizing energy into a congestion game All mobile devices inthis method are participants in the game e methodchooses a server to unload the computation tasks to optimizeQoS levels and save energy which can optimize the systemand save energy Wang et al [84] proposed a MapReduce-based multitask scheduling algorithm to achieve the ob-jective of energy saving is model is a two-layer modelwhich considers the impact of server performance changeson energy consumption and the limitation of networkbandwidth In addition a local search operator is designedbased on which a two-layer genetic algorithm is proposede algorithm can schedule tens of thousands of tasks incloud and achieve large-scale optimization Yanggratoke

et al [85] proposed a general generic gossip protocol aimingat allocating resources in cloud environment An instanti-ation of this protocol was developed to enable server con-solidation to allocate resources to more servers to meetchanging load patterns

25 LoadBalancing Load balancing can help save energy bymanaging the number of servers and allocating resources

251 Cloud Computing Paya and Marinescu [86] intro-duced an operation model that balances cloud computingload and expands applications to save energy e principleof this model is to define an operating system e systemshould make servers run in the system as many as possible

Table 5 Work summary of scientific workflow execution in cloud computing

Problems Solutions Literatures Advantages

VM deployment A resource allocation method namedEnRealan [76] Performs scientific workflows based on energy perception

across cloud platforms

Workflowscheduling

A scheduling method based on energyperception [77]

Achieves a high parallelism without huge energyconsumption and minimizes the total consumption of

energy and execution time of workflows

A workflow scheduling method with severalobjects and hybrid particle swarm

optimization algorithm[78]

Makes the processors work at any voltage level minimizesthe energy consumption in the process of workflowscheduling and studies the scheduling problem of

workflows on heterogeneous systemsA scheduling algorithm based on various

applications [79] Enables service providers to gain high profits and reducesuser overhead at the same time

Cost reduction A scheduling algorithm based on energyperception [80 81] Minimizes the cost of performing workflows while

meeting the time constraintEffectiveimplementation

A flexible resource allocation and jobscheduling mechanism [82] Implements scientific workflows within prescribed

budgets and deadlines

Table 4 Work summary of resource allocation and management in IoT

Problems Solutions Literatures Advantages

Resourcemanagement

A distributed cloud network framework [66]Replaces the centralized architecture with a distributed cloud

architecture solves the defects of the centralized cloudarchitecture and brings people better experience

A MILP model [68] Assigns the requirement of servicesrsquo resource to heterogeneousnetwork equipment interface

A framework for communication used in5G [69]

Transforms the resource allocation problem into a power andchannel allocation problem minimizes the total energy

consumption and improves QoS levelsA low complexity channel allocation

algorithm [70] Improves throughput of the network and allocates finiteresources to a large group of users

A resource allocation framework withseveral bands under cognitive 5G IoT [71] Manages resources more flexibly and reduces energy

consumption than common single-band approach

A protocol which is distributed andoptimal to allocate resources [72]

Has excellent adaptability when changing topology of networkand dynamically manages resources in the heterogeneous IoT

environmentA multilevel allocating resourcesalgorithm for IoT communication [73] Has fast data processing rate and very low latency in both

saturated and nonsaturated environment

A new algorithm to allocate bandwidthdynamically [74]

Reduces the error of signal reconstruction under the samebandwidth and makes the bandwidth allocation of IoT more

reasonable

A reinforcement learning mechanism [75]Effectively solves the resource allocation problems caused by themismatch of service quality and complex service providing

condition in the IoT

Complexity 11

When no tasks are being performed the system should adjustservers to sleep thus energy consumption can be reducedJustafort et al [87] mainly studied the problem of workloaddistribution across cloud computing environment and pro-posed a method to solve the problem of the VM layout So thefootprint of carbon can be effectively reduced Panwar andMallick [88] proposed an algorithm to dynamically manage theload and effectively distribute the total incoming requestsbetween VMs rough efficient and uniform utilization ofresources this algorithm can achieve uniform distribution ofload between servers Yang et al [89] proposed a powermanagement mechanism to balance the load e system canmonitor VMs and dynamically allocate the resources Yanget al [90] proposed an optimization system to better allocateresources dynamically which can balance the load of VMsrunning on multiple physical machines Under this systemVMs can be migrated automatically to adjust high and lowloads without interrupting services Yang et al [89 90] manageVMs to achieve load balancing ey allocate resources dy-namically to migrate VMs which can balance workloads ondifferent physical machines e difference is that Yanget al[89] integrate a dynamic resource allocation approachwithOpenNebulaWhile Yang et al [90] focuse on avoiding serviceoutages during VM migration

Table 6 shows a summary of the abovementioned workse problems in Table 6 are from the load balancing problemServer management is about the control of the number ofservers running in the system Workload management is therational allocation of workload or tasks VMmanagement refersto configuringVMresources andmigratingVMs to adjust loads

252 Fog Computing Xu et al [91] proposed a methodcalled ldquoDRAMrdquo to dynamically allocate resources in fogcomputing environment which can avoid both too highand too low loads e method first analyzes the loadbalance of different kinds of computing nodes And then itdesigns a fog resource allocation method to achieve loadbalance which allocates resources statically and migratesservices dynamically Oueis et al [92] studied the loadbalance problem in fog computing A custom fog clus-tering algorithm is proposed to solve the problem In theproblem several users need to offload computations andall of their demands need to be handled by local com-puting cluster

253 IoT Wang et al [93] established architecture of theenergy saving targeted system in industrial IoT In additionin order to predict sleep intervals they developed a sleepscheduling and a wake protocol which provide a better wayfor energy saving

3 Conclusion

is paper did a comprehensive study of QoS optimi-zation and energy saving in cloud computing edgecomputing fog computing and IoT models We sum-marized five main problems and analyzed their solutionsproposed by existing works By conducting this surveywe aim to help readers have a deeper understanding onthe concepts of different computing models and thetechniques of QoS optimization and energy saving inthese models

e investigated papers focus on issues about ensuringQoS and reducing SLA violations and resource manage-ment In the case of QoS assurance and SLA violationsreduction the main solution of QoS assurance is efficientVM management is solution can meet customersrsquo re-quirements through reasonable scheduling and integrationof VMs Most of resource management techniques are re-alized by reasonable scheduling of resources which canreduce the waste of VMs servers and traffic

Disclosure

is manuscript is an extension of A Survey of QoS Opti-mization and Energy Saving in Cloud Edge and IoT in e9th EAI International Conference on Cloud Computing

Conflicts of Interest

e authors declare that there are no conflicts of interestregarding the publication of this paper

Acknowledgments

is work was supported by the National Natural ScienceFoundation of China (Grant no 61702274) Natural ScienceFoundation of Jiangsu Province (Grant no BK20170958)and PAPD

Table 6 Work summary of load balancing in cloud computing

Problems Solutions Literatures AdvantagesServermanagement

An operation model that can balance cloudcomputing load and expand application [86] Saves energy by managing the number of servers running in

the system

Workloadmanagement

A hybrid approach [87] Reduces the footprint of carbon and allocates workload acrosscloud computing

An algorithm to dynamically manage theload [88] Manages the load evens the load distribution between

servers and allocates tasks between VMs

VMmanagement

A green power administration mechanism [89] Monitors and jointly allocates VM resources

An optimization system [90]Migrates VMs to adjust high and low loads without

interrupting services and balances the load of VMs runningon multiple physical machines

12 Complexity

References

[1] Q Zhang L Cheng and R Boutaba ldquoCloud computingstate-of-the-art and research challengesrdquo Journal of InternetServices and Applications vol 1 no 1 pp 7ndash18 2010

[2] W Shi J Cao Q Zhang Y Li and L Xu ldquoEdge computingvision and challengesrdquo IEEE Internet of Aings Journal vol 3no 5 pp 637ndash646 2016

[3] M Iorga L Feldman R Barton M J Martin N S Gorenand C Mahmoudi ldquoFog computing conceptual modelrdquo TechRep Recommendations of the National Institute of Standardsand Technology Gaithersburg MD USA 2018

[4] Nebbiolo ldquoFog vs edge computingrdquo Tech Rep NebbioloTechnologies Inc Milpitas CA USA 2018

[5] C T Do N H Tran C Pham M G R Alam J H Son andC S Hong ldquoA proximal algorithm for joint resource allo-cation andminimizing carbon footprint in geo-distributed fogcomputingrdquo in Proceedings of the 2015 International Con-ference on Information Networking (ICOIN) pp 324ndash329IEEE Siem Reap Cambodia January 2015

[6] S Vashi J Ram J Modi S Verma and C Prakash ldquoInternetof things (IoT) a vision architectural elements and securityissuesrdquo in Proceedings of the 2017 International Conference onI-SMAC (IoT in Social Mobile Analytics and Cloud)(I-SMAC) pp 492ndash496 IEEE Coimbatore India February2017

[7] M Mazzucco D Dyachuk and R Deters ldquoMaximizing cloudprovidersrsquo revenues via energy aware allocation policiesrdquo inProceedings of the 2010 IEEE 3rd International Conference onCloud Computing pp 131ndash138 IEEE Miami FL USA July2010

[8] Q He R Zhou X Zhang et al ldquoKeyword search for buildingservice-based systemsrdquo IEEE Transactions on Software En-gineering vol 43 no 7 pp 658ndash674 2016

[9] L Sun H Dong O K Hussain F K Hussain and A X LiuldquoA framework of cloud service selection with criteria inter-actionsrdquo Future Generation Computer Systems vol 94pp 749ndash764 2019

[10] M Mazzucco and D Dyachuk ldquoOptimizing cloud providersrevenues via energy efficient server allocationrdquo SustainableComputing Informatics and Systems vol 2 no 1 pp 1ndash122012

[11] Q He J Han F Chen et al ldquoQos-aware service selection forcustomisable multi-tenant service-based systems maturityand approachesrdquo in Proceedings of the 2015 IEEE 8th Inter-national Conference on Cloud Computing pp 237ndash244 IEEENew York NY USA July 2015

[12] L Sun J Ma H Wang Y Zhang and J Yong ldquoCloud servicedescription model an extension of usdl for cloud servicesrdquoIEEE Transactions on Services Computing vol 11 no 2pp 354ndash368 2015

[13] Y Wang Q He D Ye and Y Yang ldquoFormulating criticality-based cost-effective fault tolerance strategies for multi-tenantservice-based systemsrdquo IEEE Transactions on Software En-gineering vol 44 no 3 pp 291ndash307 2017

[14] S Mustafa K Bilal S U R Malik and S A Madani ldquoSla-aware energy efficient resource management for cloud en-vironmentsrdquo IEEE Access vol 6 pp 15004ndash15020 2018

[15] J Bi H Yuan M Tie and W Tan ldquoSla-based optimisation ofvirtualised resource for multi-tier web applications in clouddata centresrdquo Enterprise Information Systems vol 9 no 7pp 743ndash767 2015

[16] S Singh I Chana and R Buyya ldquoStar sla-aware autonomicmanagement of cloud resourcesrdquo IEEE Transactions on CloudComputing p 1 2017

[17] A Beloglazov and R Buyya ldquoEnergy efficient resourcemanagement in virtualized cloud data centersrdquo in Proceedingsof the 2010 10th IEEEACM International Conference onCluster Cloud and Grid Computing pp 826ndash831 IEEEComputer Society Melbourne Australia May 2010

[18] M Guazzone C Anglano and M Canonico ldquoEnergy-effi-cient resource management for cloud computing infra-structuresrdquo in Proceedings of the 2011 IEEE AirdInternational Conference on Cloud Computing Technology andScience pp 424ndash431 IEEE Athens Greece November 2011

[19] Y Sun J White and S Eade ldquoA model-based system toautomate cloud resource allocation and optimizationrdquo inProceedings of the International Conference on Model DrivenEngineering Languages and Systems pp 18ndash34 SpringerValencia Spain October 2014

[20] G Siddesh and K Srinivasa ldquoSla-driven dynamic resourceallocation on cloudsrdquo in Proceedings of the InternationalConference on Advanced Computing Networking and Securitypp 9ndash18 Springer Surathkal India December 2011

[21] S K Garg S K Gopalaiyengar and R Buyya ldquoSla-basedresource provisioning for heterogeneous workloads in avirtualized cloud datacenterrdquo in Proceedings of the Interna-tional Conference on Algorithms and Architectures for ParallelProcessing pp 371ndash384 Springer Melbourne AustraliaOctober 2011

[22] J Bi Z Zhu and H Yuan ldquoSla-aware dynamic resourceprovisioning for profit maximization in shared cloud datacentersrdquo in Proceedings of the International Conference onHigh Performance Networking Computing and Communi-cation Systems pp 366ndash372 Springer Singapore May 2011

[23] L Qi Y Chen Y Yuan S Fu X Zhang and X Xu ldquoA Qos-aware virtual machine scheduling method for energy con-servation in cloud-based cyber-physical systemsrdquo WorldWide Web vol 4 no 3 pp 1ndash23 2019

[24] A Beloglazov J Abawajy and R Buyya ldquoEnergy-aware re-source allocation heuristics for efficient management of datacenters for cloud computingrdquo Future Generation ComputerSystems vol 28 no 5 pp 755ndash768 2012

[25] L Qi R Wang C Hu S Li Q He and X Xu ldquoTime-awaredistributed service recommendation with privacy-preserva-tionrdquo Information Sciences vol 480 pp 354ndash364 2019

[26] Z Zhai B Cheng Y Tian J Chen L Zhao and M Niu ldquoAdata-driven service creation approach for end-usersrdquo IEEEAccess vol 4 pp 9923ndash9940 2016

[27] L Gu D Zeng S Guo A Barnawi and Y Xiang ldquoCostefficient resource management in fog computing supportedmedical cyber-physical systemrdquo IEEE Transactions onEmerging Topics in Computing vol 5 no 1 pp 108ndash119 2015

[28] L Ni J Zhang C Jiang C Yan and K Yu ldquoResource al-location strategy in fog computing based on priced timed petrinetsrdquo Ieee Internet of Aings Journal vol 4 no 5pp 1216ndash1228 2017

[29] L Wei T Yang F C Delicato et al ldquoOn enabling sustainableedge computing with renewable energy resourcesrdquo IEEECommunications Magazine vol 56 no 5 pp 94ndash101 2018

[30] P Lai Q He G Cui et al ldquoEdge user allocation with dynamicquality of servicerdquo in Proceedings of the International Con-ference on Service-Oriented Computing pp 86ndash101 SpringerToulouse France October 2019

[31] X Xu Q Liu X Zhang J Zhang L Qi and W Dou ldquoAblockchain-powered crowdsourcing method with privacy

Complexity 13

preservation in mobile environmentrdquo IEEE Transactions onComputational Social Systems vol 6 no 6 pp 1407ndash1419 2019

[32] O Rolik E Zharikov and S Telenyk ldquoMicrocloud-basedarchitecture of management system for IoT infrastructuresrdquoin Proceedings of the 2016 Aird International Scientific-Practical Conference Problems of Infocommunications Scienceand Technology (PIC SampT) pp 149ndash151 IEEE KharkivUkaraine October 2016

[33] W He S Guo Y Liang R Ma X Qiu and L Shi ldquoQos-awareand resource-efficient dynamic slicingmechanism for internetof thingsrdquo Computers Materials amp Continua vol 61 no 3pp 1345ndash1364 2019

[34] J Yao and N Ansari ldquoQos-aware fog resource provisioningand mobile device power control in IoT networksrdquo IEEETransactions on Network and Service Management vol 16no 1 pp 167ndash175 2018

[35] W Wang Y Jiang and W Wu ldquoMultiagent-based resourceallocation for energy minimization in cloud computing sys-temsrdquo IEEE Transactions on Systems Man and CyberneticsSystems vol 47 no 2 pp 1ndash16 2016

[36] K Krishnajyothi ldquoParallel data processing for effective dy-namic resource allocation in the cloudrdquo International Journalof Computer Applications vol 70 no 22 pp 1ndash4 2013

[37] M M Hassan B Song M S Hossain and A Alamri ldquoEf-ficient resource scheduling for big data processing in cloudplatformrdquo in Proceedings of the International Conference onInternet and Distributed Computing Systems pp 51ndash63Springer Calabria Italy September 2014

[38] C-M Wu R-S Chang and H-Y Chan ldquoA green energy-efficient scheduling algorithm using the dvfs technique forcloud datacentersrdquo Future Generation Computer Systemsvol 37 no 7 pp 141ndash147 2014

[39] Z Wang and X Su ldquoDynamically hierarchical resource-al-location algorithm in cloud computing environmentrdquo AeJournal of Supercomputing vol 71 no 7 pp 2748ndash2766 2015

[40] W-Y Lin G-Y Lin and H-Y Wei ldquoDynamic auctionmechanism for cloud resource allocationrdquo in Proceedings ofthe 2010 10th IEEEACM International Conference on ClusterCloud and Grid Computing pp 591-592 IEEE ComputerSociety Victoria Australia May 2010

[41] Y O Yazir C Matthews R Farahbod et al ldquoDynamic re-source allocation in computing clouds using distributedmultiple criteria decision analysisrdquo in Proceedings of the 2010IEEE 3rd International Conference on Cloud Computingpp 91ndash98 IEEE Washington DC USA July 2010

[42] W Lin J Z Wang C Liang and D Qi ldquoA threshold-baseddynamic resource allocation scheme for cloud computingrdquoProcedia Engineering vol 23 no 5 pp 695ndash703 2011

[43] X Xu S Fu L Qi et al ldquoAn IoT-oriented data placementmethod with privacy preservation in cloud environmentrdquoJournal of Network and Computer Applications vol 124pp 148ndash157 2018

[44] J Bokyun P Md Jalil L Daeho and S Doug Young ldquoEf-ficient computation offloading in mobile cloud computing forvideo streaming over 5 grdquo Computers Materials amp Continuavol 61 no 2 pp 439ndash463 2019

[45] H Sarbazi-Azad and A Y Zomaya ldquoEnergy-efficient resourceutilization in cloud computingrdquo in Large Scale Network-Centric Distributed Systems pp 377ndash408 Wiley-IEEE PressHoboken NJ USA 2014

[46] C-H Hsu K D Slagter S-C Chen and Y-C ChungldquoOptimizing energy consumption with task consolidation incloudsrdquo Information Sciences vol 258 no 3 pp 452ndash4622014

[47] C-H Hsu S-C Chen C-C Lee et al ldquoEnergy-aware taskconsolidation technique for cloud computingrdquo in Proceedingsof the 2011 IEEE Aird International Conference on CloudComputing Technology and Science pp 115ndash121 IEEE Ath-ens Greece 2011

[48] S K Panda and P K Jana ldquoAn efficient task consolidationalgorithm for cloud computing systemsrdquo in Proceedings of theInternational Conference on Distributed Computing and In-ternet Technology pp 61ndash74 Springer Bhubaneswar IndiaJanuary 2016

[49] L Yin J Luo and H Luo ldquoTasks scheduling and resourceallocation in fog computing based on containers for smartmanufacturingrdquo IEEE Transactions on Industrial Informaticsvol 14 no 10 pp 4712ndash4721 2018

[50] M Aazam and E-N Huh ldquoDynamic resource provisioningthrough fog micro datacenterrdquo in Proceedings of the 2015IEEE International Conference on Pervasive Computing andCommunication Workshops (PerCom Workshops) pp 105ndash110 IEEE St Louis MO USA March 2015

[51] B Jia H Hu Y Zeng T Xu and Y Yang ldquoDouble-matchingresource allocation strategy in fog computing networks basedon cost efficiencyrdquo Journal of Communications and Networksvol 20 no 3 pp 237ndash246 2018

[52] H Zhang Y Xiao S Bu D Niyato F R Yu and Z HanldquoComputing resource allocation in three-tier IoT fog net-works a joint optimization approach combining stackelberggame and matchingrdquo IEEE Internet of Aings Journal vol 4no 5 pp 1204ndash1215 2017

[53] J Tan T-H Chang and T Q Quelc ldquoMinimum energyresource allocation in fog radio access network with fronthauland latency constraintsrdquo in Proceedings of the 2018 IEEE 19thInternational Workshop on Signal Processing Advances inWireless Communications (SPAWC) pp 1ndash5 IEEE KalamataGreece June 2018

[54] D R de Vasconcelos R M de Castro Andrade andJ N de Souza ldquoSmart shadowndashan autonomous availabilitycomputation resource allocation platform for internet ofthings in the fog computing environmentrdquo in Proceedings of the2015 International Conference on Distributed Computing inSensor Systems pp 216-217 IEEE Fortaleza Brazil June 2015

[55] M Aazam M St-Hilaire C-H Lung and I LambadarisldquoPre-fog IoT trace based probabilistic resource estimation atfogrdquo in Proceedings of the 2016 13th IEEE Annual ConsumerCommunications amp Networking Conference (CCNC) pp 12ndash17 IEEE Las Vegas NV USA January 2016

[56] N D Tung L L Bao and B Vijay ldquoPrice-based resourceallocation for edge computing a market equilibrium ap-proachrdquo IEEE Transactions on Cloud Computing p 1 2018

[57] X Xu C He Z Xu L Qi S Wan and M Z A BhuiyanldquoJoint optimization of offloading utility and privacy for edgecomputing enabled IoTrdquo IEEE Internet of Aings Journal2019

[58] X Xu Y Chen X Zhang Q Liu X Liu and L Qi ldquoAblockchain-based computation offloading method for edgecomputing in 5G networksrdquo Software Practice and Experi-ence 2019

[59] X Xu Y Li T Huang et al ldquoAn energy-aware computationoffloading method for smart edge computing in wirelessmetropolitan area networksrdquo Journal of Network and Com-puter Applications vol 133 pp 75ndash85 2019

[60] X Xu Y Xue L Qi et al ldquoAn edge computing-enabledcomputation offloading method with privacy preservation forinternet of connected vehiclesrdquo Future Generation ComputerSystems vol 96 pp 89ndash100 2019

14 Complexity

[61] G Yeting L Fang X Nong and C Zhengguo ldquoTask-basedresource allocation bid in edge computing micro datacenterrdquoComputers Materials amp Continua vol 61 no 2 pp 777ndash7922019

[62] X Chen L Jiao W Li and X Fu ldquoEfficient multi-usercomputation offloading for mobile-edge cloud computingrdquoIEEEACM Transactions on Networking vol 24 no 5pp 2795ndash2808 2016

[63] B Gao L He X Lu C Chang K Li and K Li ldquoDevelopingenergy-aware task allocation schemes in cloud-assisted mo-bile workflowsrdquo in Proceedings of the 2015 IEEE InternationalConference on Computer and Information Technology Ubiq-uitous Computing and Communications Dependable Auto-nomic and Secure Computing Pervasive Intelligence andComputing pp 1266ndash1273 IEEE Liverpool UK October2015

[64] X Xu X Zhang H Gao Y Xue L Qi andW Dou ldquoBecomeblockchain-enabled computation offloading for IoT in mobileedge computingrdquo IEEE Transactions on Industrial Infor-matics 2019

[65] W Yifei W Zhaoying G Da and Y F Richard ldquoDeepq-learning based computation offloading strategy for mobileedge computingrdquo Computers Materials amp Continua vol 59no 1 pp 89ndash104 2019

[66] M Barcelo A Correa J Llorca A M Tulino J L Vicarioand A Morell ldquoIoT-cloud service optimization in nextgeneration smart environmentsrdquo IEEE Journal on SelectedAreas in Communications vol 34 no 12 pp 4077ndash40902016

[67] B Cheng M Wang S Zhao Z Zhai D Zhu and J ChenldquoSituation-aware dynamic service coordination in an IoTenvironmentrdquo IEEEACM Transactions On Networkingvol 25 no 4 pp 2082ndash2095 2017

[68] V Angelakis I Avgouleas N Pappas and D Yuan ldquoFlexibleallocation of heterogeneous resources to services on an IoTdevicerdquo in Proceedings of the 2015 IEEE Conference onComputer Communications Workshops (INFOCOMWKSHPS) pp 99-100 IEEE Hong Kong China April 2015

[69] S Li Q Ni Y Sun G Min and S Al-Rubaye ldquoEnergy-ef-ficient resource allocation for industrial cyber-physical IoTsystems in 5g erardquo IEEE Transactions on Industrial Infor-matics vol 14 no 6 pp 2618ndash2628 2018

[70] X Liu Z Qin Y Gao and J A McCann ldquoResource allocationin wireless powered IoT networksrdquo IEEE Internet of AingsJournal vol 6 no 3 pp 4935ndash4945 2019

[71] W Ejaz and M Ibnkahla ldquoMulti-band spectrum sensing andresource allocation for IoT in cognitive 5g networksrdquo IEEEInternet of Aings Journal vol 5 no 1 pp 150ndash163 2017

[72] G Colistra V Pilloni and L Atzori ldquoTask allocation in groupof nodes in the IoT a consensus approachrdquo in Proceedings ofthe 2014 IEEE International Conference on Communications(ICC) pp 3848ndash3853 IEEE Sydney Australia June 2014

[73] J Li Q Sun and G Fan ldquoResource allocation for multiclassservice in IoTuplink communicationsrdquo inProceedings of the 20163rd International Conference on Systems and Informatics (ICSAI)pp 777ndash781 IEEE Shanghai China November 2016

[74] L I Zheng and K H Liu ldquoDynamic bandwidth resourceallocation algorithm in internet of things and its applicationrdquoComputer Engineering vol 38 no 17 pp 16ndash19 2012

[75] K Gai and M Qiu ldquoOptimal resource allocation using re-inforcement learning for IoT content-centric servicesrdquo Ap-plied Soft Computing vol 70 pp 12ndash21 2018

[76] X Xu W Dou X Zhang and J Chen ldquoEnreal an energy-aware resource allocation method for scientific workflow

executions in cloud environmentrdquo IEEE Transactions onCloud Computing vol 4 no 2 pp 166ndash179 2016

[77] K Bousselmi Z Brahmi and M M Gammoudi ldquoEnergyefficient partitioning and scheduling approach for scientificworkflows in the cloudrdquo in Proceedings of the 2016 IEEEInternational Conference on Services Computing (SCC)pp 146ndash154 IEEE San Francisco CA USA June 2016

[78] Y Sonia C Rachid K Hubert and G Bertrand ldquoMulti-objective approach for energy-aware workflow scheduling incloud computing environmentsrdquo Ae Scientific World Jour-nal vol 2013 Article ID 350934 13 pages 2013

[79] F Cao Efficient Scientific Workflow Scheduling in CloudEnvironment Southern Illinois University Carbondale ILUSA 2014

[80] Z Li J Ge H Hu W Song H Hu and B Luo ldquoCost andenergy aware scheduling algorithm for scientific workflowswith deadline constraint in cloudsrdquo IEEE Transactions onServices Computing vol 11 no 4 pp 713ndash726 2015

[81] M Khaleel and M M Zhu ldquoEnergy-aware job managementapproaches for workflow in cloudrdquo in Proceedings of the 2015IEEE International Conference on Cluster Computingpp 506-507 IEEE Chicago IL USA September 2015

[82] J Shi J Luo F Dong and J Zhang ldquoA budget and deadlineaware scientific workflow resource provisioning and sched-uling mechanism for cloudrdquo in Proceedings of the 2014 IEEE18th International Conference on Computer Supported Co-operative Work in Design (CSCWD) pp 672ndash677 IEEEHsinchu Taiwan January 2014

[83] Y Ge Y Zhang Q Qiu and Y-H Lu ldquoA game theoreticresource allocation for overall energy minimization in mobilecloud computing systemrdquo in Proceedings of the 2012ACMIEEE International Symposium on Low Power Electronicsand Design pp 279ndash284 ACM Redondo Beach CA USAAugust 2012

[84] X Wang Y Wang and C Yue ldquoAn energy-aware bi-leveloptimization model for multi-job scheduling problems undercloud computingrdquo Soft Computing vol 20 no 1 pp 303ndash3172016

[85] R Yanggratoke F Wuhib and R Stadler ldquoGossip-basedresource allocation for green computing in large cloudsrdquo inProceedings of the 2011 7th International Conference onNetwork and Service Management pp 1ndash9 IEEE ParisFrance October 2011

[86] A Paya and D C Marinescu ldquoEnergy-aware load balancingand application scaling for the cloud ecosystemrdquo IEEETransactions on Cloud Computing vol 5 no 1 pp 15ndash272015

[87] V D Justafort R Beaubrun and S Pierre ldquoA hybrid ap-proach for optimizing carbon footprint in intercloud envi-ronmentrdquo IEEE Transactions on Services Computing vol 12no 2 pp 186ndash198 2016

[88] R Panwar and B Mallick ldquoLoad balancing in cloud com-puting using dynamic load management algorithmrdquo inProceedings of the 2015 International Conference on GreenComputing and Internet of Aings (ICGCIoT) pp 773ndash778IEEE Noida India October 2015

[89] C-T Yang K-C Wang H-Y Cheng C-T Kuo andW C C Chu ldquoGreen power management with dynamicresource allocation for cloud virtual machinesrdquo in Proceedingsof the 2011 IEEE International Conference on High Perfor-mance Computing and Communications pp 726ndash733 IEEEAlberta Canada September 2011

[90] C-T Yang H-Y Cheng and K-L Huang ldquoA dynamicresource allocation model for virtual machine management

Complexity 15

on cloudrdquo in Proceedings of the International Conference onGrid and Distributed Computing pp 581ndash590 SpringerGangneug Korea December 2011

[91] X Xu F Shucun C Qing et al ldquoDynamic resource allocationfor load balancing in fog environmentrdquo Wireless Commu-nications amp Mobile Computing vol 2018 no 2 Article ID6421607 15 pages 2018

[92] J Oueis E C Strinati and S Barbarossa ldquoe fog balancingload distribution for small cell cloud computingrdquo in Pro-ceedings of the 2015 IEEE 81st Vehicular Technology Confer-ence (VTC Spring) pp 1ndash6 IEEE Glasgow UK May 2015

[93] K Wang Y Wang Y Sun S Guo and J Wu ldquoGreen in-dustrial internet of things architecture an energy-efficientperspectiverdquo IEEE Communications Magazine vol 54 no 12pp 48ndash54 2016

16 Complexity

Tabl

e1

Worksummaryof

QoS

guaranteeing

orSL

Aassurancein

clou

dcompu

ting

Subp

roblem

sSolutio

nsLiteratures

Advantages

Server

managem

ent

Apo

licyof

dynamicallocatio

nof

poweringserverss

witches

[71

0]Maxim

izes

benefitsim

proves

QoSa

ndminim

izes

power

consum

ption

Workloads

consolidation

Atechniqu

eforconsolidatingworkloads

[14]

Achievesenergy

saving

sby

usingfewestserverswhile

redu

cing

SLA

violations

VM

managem

ent

Anarchitecturethat

canadministrate

itselfa

ndamixed

queuingmod

el[15]

Makes

morerevenu

esandmeets

different

requ

irem

ents

ofcustom

ersanddecidesthe

numberof

VMsforeach

layerof

avirtuala

pplication

Asystem

tointegrateanddynamically

redistribu

teVMs

[17]

Achievesthe

goalofsaving

energy

throughVM

integrationandprovidesah

ighQoS

levelat

thesametim

e

AQoS-awareVM

schedu

lingstrategy

named

QVMS

[23]

Effectiv

elymanages

resourcesin

thenetworkph

ysical

system

toredu

cetheenergy

consum

ptionandim

proves

QoS

AVM

integrationmetho

dwith

severaltargets

[24]

Savesenergy

andredu

cesSL

Aviolations

byapplying

different

strategies

todifferent

load

states

oftheho

st

Self-managem

ent

Atechno

logy

named

STAR

[16]

Redu

cesSL

Aviolations

andim

proves

paym

ente

fficiency

ofclou

dservices

Adynamic

resource

managem

entsystem

[18]

Self-manages

theresourceso

fcloud

infrastructurestoprovidea

ppropriateQoS

andfitsthe

changing

workloads

dynamically

Resource

managem

ent

Amod

elthat

canrealizetheindepend

entallocatio

nof

resources

[19]

Obtains

theop

timal

resource

confi

guratio

nmeets

theQoS

requ

irem

entsand

provides

econ

omical

clou

dresources

Adynamic

resource

allocatio

nstrategy

[20

21]

Redu

cesSL

Aandmaxim

izes

revenu

esandresource

utilizatio

non

theclou

d

Amixed

queuemod

el[22]

Reason

ably

confi

gurestheresourcesin

theclou

ddata

centerimproves

thesystem

performancereduces

theadditio

nalcosto

fusin

gresourcesmeetstherequ

ired

QoSand

provides

virtualresou

rces

toeach

layerof

virtuala

pplicationservices

Service

managem

ent

Aun

ified

semantic

mod

elthat

candescribe

clou

dservice

[12]

Improves

theability

ofmod

elon

servicerank

ingandenriches

thelang

uage

expressio

n

Arecommendatio

nservicestrategy

[25]

Emph

asizes

theinflu

ence

oftim

efactorson

QoS

andim

proves

thetradition

allocatio

n-sensitive

hash

techno

logy

toprotectusersrsquoprivacy

Aclou

dserviceselectionmetho

dusingfuzzymeasure

and

Cho

quet

integral

andafram

eworkbasedon

priority

[9]

Selectsservicewhenhistorical

inform

ationisinsufficientto

determ

inethecriteria

relatio

nships

andweigh

ts

reeserviceselectionmetho

dsthat

supp

ortQ

oSandcan

combine

multitenants

ervice-based

system

s[11]

Achievesthree

degreeso

fmultitenantm

aturity

which

ismoreeffi

cientthanthetradition

alsin

gle-user

approach

Afault-tolerant

strategy

basedon

multitenants

ervice

criticality

[13]

Guaranteesthequ

ality

ofthemultitenant-basedservicesystem

Aservice-basedsystem

supp

ortin

gkeyw

ordsearch

[8]

Effectiv

elyhelpssystem

engineerswho

areno

tfam

iliar

with

service-oriented

architecture

techno

logy

tobu

ildservice-oriented

system

s

6 Complexity

merges tasks in virtual clusters Once a task migrationhappens the energy cost model will take into account thenetwork latency Sarbazi-Azad and Zomaya [45] and Hsuet al [46] both maximize the benefit of cloud resources byusing task merging techniques Sarbazi-Azad and Zomaya[45] uses a greedy algorithm called MaxUtil While Hsuet al [46] takes into account the network latency associatedwith task migration So in [46] a 17 improvement isachieved over MaxUtil

E0m Vi( 1113857 1113944m

t0Et Vi( 1113857 (6)

E0m VCk( 1113857 1113944n

i0E0m Vi( 1113857 (7)

where Et is the energy consumption in unit time and n is thenumber of VMs in the cluster

Hsu et al [47] proposed a task integration technologybased on the energy perception According to the charac-teristics of most cloud systems the principle of using 70CPU is proposed to administrate job integration amongvirtual clusters is technology is very effective in reducingthe amount of energy consumed in cloud systems bymerging tasks Panda and Jana [48] proposed an algorithmwith several criteria to combine tasks e algorithm notonly considers the time needed for processing jobs but alsoconsiders the utilization rate of VMs And the algorithm ismore energy efficient because it takes into account not onlythe processing time but also the utilization rate of VMsWang and Su [39] proposed a resource allocation algorithmto deal with wide range of communication between nodes incloud environment is algorithm uses recognition tech-nology to dynamically distribute jobs and nodes accordingto computing ability and factors of storage And it can re-duce the traffic when allocating resources because it usesdynamic hierarchy Lin et al [40] proposed a dynamicauction approach for resource allocation e approach canensure that even if there are many users and resourcesproviders will have reasonable profits and computing re-sources will be allocated correctly Yazir et al [41] proposeda new method to manage resources dynamically and au-tonomously Firstly resource management is split into jobsand each job is executed by autonomous nodes Secondautonomous nodes use the method called PROMETHEE toconfigure resources Krishnajyothi [36] proposed a frame-work which can implement parallel task processing to solvethe problem of low efficiency when submitting large tasksCompared with the static framework this framework candynamically allocate VMs thus reducing costs and the timeof processing tasks Lin et al [42] proposed a method toallocate resources dynamically by using thresholds Becausethis method uses the threshold value it can optimize thereallocation of resources improve the usage of resourcesand reduce the cost Xu et al [43] proposed a data placementstrategy named IDP for the data generated by IoTs devices toachieve reasonable data placement In this way the privacyof these data can be protected while resources are allocatedreasonably Jo et al [44] proposed a computing offload

framework under 5G network e framework transfers thecomputing burden to the cloud thus reducing the com-puting load of clients and the communication cost

Table 2 shows a summary of the abovementioned workse problem of resource allocation and management incloud computing can be divided into problems in Table 2VM management is about a reasonable configuration ofVMs Resource allocation represents the dynamic andflexible allocation of resources Task integration refers tocombining tasks to save energy and improve efficiency

222 Fog Computing Yin et al [49] established a newmodel of scheduling jobs which applies containers In orderto make sure that jobs can be finished on time a jobscheduling algorithm is developed e algorithm can alsooptimize the number of tasks that can be performed togetheron the nodes in fog computing And this paper proposes aredistribution mechanism to shorten the delay of tasksese methods are very effective in reducing task delaysAazam and Huh [50] established a framework to admin-istrate resources effectively in the mode of fog computingConsidering that there are various types of objects anddevices the connection between them may be volatile So amethod for predicting and administrating resources isproposed e method considers that any object or devicecan quit using resources at anytime Cuong et al [5] studiedthe allocating resources jointly problem and the problem ofcarbon footprint minimization in fog data center Formula(8) [5] is used to denote the energy consumption of serversIn addition a distributed algorithm is proposed to solve theproblem of wide range optimization

P(y) C middot Pidle + Ppeak minus Pidle1113872 1113873 middot y middot κ (8)

where P(y) represents the power supply required by theservers in a data center y represents the video stream κdenotes a conversion factor that converts the video streaminto workload C represents the data centerrsquos load capacityand Pidle and Ppeak respectively represent the idle power andpeak power of the servers

Jia et al [51] studied the problem of computing resourceallocation in fog computing network with three levelsFirstly the problem of resource allocation is transformedinto a bilateral matching optimal problem And then a bi-matching approach is proposed for this problem which canimprove the performance of the system and obtain highercost efficiency Zhang et al [52] proposed a framework forjoint optimization under fog computing to allocate fognodesrsquo finite computing resources e framework canachieve the best allocation and effectively improve thenetworksrsquo performance Tan et al [53] presented a method toallocate computing and communication resources emethod transfers computing jobs to remote cloud and nodesand simplifies edge nodesrsquo computing and computing en-ergy Vasconcelos et al [54] developed a platform to allocateresources accessible to client devices in fog computingenvironment allocating the resources of devices near thehost to meet the applications needs for rapid response tocomputing resources Aazam et al [55] presented a method

Complexity 7

to estimate and manage resource in fog computing emethod is based on the fluctuation of customer abandon-ment probability type and price of service and so on

Table 3 shows a summary of the abovementioned workse problems in Table 3 are also derived from resourceallocation and management problems Task allocationrepresents the scheduling and redistribution of tasks Re-source allocation is still about the dynamic and flexibleallocation of resources Low latency refers to taking shorttime to configure and manage resources which can improveefficiency

223 Edge Computing Tung et al [56] proposed a newframework for resource allocation based on market needse resources come from edge nodes (ENs) with limitedheterogeneous capabilities and are allocated to multiplecompeting services on the network edge Generating amarket equilibrium solution by reasonably pricing ENs canobtain the maximum utilization of marginal computingresources Xu et al [57] proposed a strategy to optimizeoffloading and privacy protection is strategy shifts tasksfirstly to improve the resource utilization of resource-limited

edge cells And then it balances QoS performance andprivacy protection to achieve joint optimization

Xu et al [58] proposed an offload strategy for edgecomputing under 5G network which uses block chaintechnologye optimal strategy is further obtained by usingthe balanced offloading method It solves the problem ofdata loss under the condition of transmission delay which iscaused by the uneven requirements of user equipments onresources Xu et al [59] proposed a computational off-loading method named EACO to reduce the energy con-sumption in smart computing models Figure 5 showsarchitecture of smart edge computing where the shortestpath is used to unload tasks EACO uses genetic algorithmsto reduce the energy consumption for operating edgecomputing nodes and improve the efficiency of performingcomplex computing tasks Xu et al [60] proposed a com-putational offloading strategy for edge computing to protectthe privacy of interconnected vehicle networks ey firstlyanalyzed privacy conflicts of tasks And then they designedthe communication route to obtain routing vehicles whichcan achieve the optimization of several objectives Yetinget al [61] proposed a unique resource allocation mechanisme mechanism takes each individual task as the basis for

Table 2 Work summary of resource allocation and management in cloud computing

Subproblems Solutions Literatures Advantages

VMmanagement

A VM allocation method based on severaldistributed agents [35] Centralizes VMs to physical machines and reduces the

overall energy costA framework which can implement parallel

task processing [36] Dynamically allocates VMs for large tasks

Resourceallocation

A general problem formula and a heuristicalgorithm with optimization parameters [37]

Dynamically allocates resources according to QoSrequirements and realizes energy saving by optimizing the

number of serversA scheduling algorithm based on thetechnology that can scale the voltage

frequency dynamically[38] Ensures the performance of executing jobs while

implementing green computing

A fuzzy pattern recognition technology [39] Reduces the traffic when allocating resources and usesdynamic hierarchy

A dynamic auction approach [40]Guarantees profits of providers and allocates resourcescorrectly when there are a large amount of users and

resources

PROMETHEE [41]Allows node agents to decompose and execute tasksautonomously to improve the flexibility of resource

allocation

A method using thresholds [42]Optimizes resource reallocation improves resource usagereduces cost and studies resource allocation strategies at the

application level

IoT-oriented data placement (IDP) method [43] Protects data privacy allocates resources reasonably andfocuses on the placement method of IoT data

A computing offload framework under 5Gnetwork [44] Reduces the computing load of clients and the

communication cost

Taskconsolidation

Two job consolidation heuristics namedldquoMaxUtilrdquo and energy-conscious task

consolidation[45] Promotes concurrent execution of multiple tasks and

improves energy efficiency

A job consolidation technique aiming atenergy saving [46 47]

Reduces the power consumption of cloud system protectsdata privacy allocates resources reasonably limits CPU

usage and merges tasks in virtual clusters

An algorithm with several criteria [48]Takes into account the job processing time and the VM

utilization rate simultaneously Dramatically reduces energyconsumption compared with state-of-the-art works

8 Complexity

resource allocation rather than for the whole service It reducesthe packet loss rate and saves energy by unloading services

224 MEC Chen et al [62] studied the problem of com-puting unloading with several users in the environment ofMEC with wireless interference which have many channels Inaddition a distributed algorithm for computing unloading isdevelopede algorithm can perform the unloading well evenwhen there are a large number of users Gao et al [63] built aquadratic binary program which is able to assign tasks inmobile cloud computing environment Two algorithms arepresented to obtain the optimal solution Both of these heuristicalgorithms can effectively solve the task assignment problemXu et al [64] proposed an offloadingmethod using block chaintechnology It can guarantee the loss of data in offloading tasksunder edge computing And it can solve the problem of re-source requests out of proportion due to the limited load of

edge computing equipment during task transfer Yifei et al [65]proposed a model-free reinforcement learning framework tosolve the problem of computational unloadingis model canbe applied to the computational unloading with time-changingcomputing requests

225 IoT Barcelo et al [66] expressed the problem ofservice allocation [67] as a mixed flow problem with min-imum cost which can be solved by LP solving this serviceallocation problem can solve the problems of unbalancednetwork load and delay of end-to-end service And it canalso figure out the problem of excessive consumption ofelectricity brought by the architecture of centralized cloudAngelakis et al [68] assigned the requirements of servicesresources to heterogeneous network interfaces of equipmentsSo more heterogeneous network interfaces can be used by alarge amount of services

Mobile devices

Mobile devices ECN

ECN

ECNECN

AP

AP

AP

APAP

Figure 5 An architecture of smart edge computing [59]

Table 3 Work summary of resource allocation and management in fog computing

Problems Solutions Literatures AdvantagesTaskallocation

A new model of scheduling jobs and aredistribution mechanism [49] Finishes jobs on time optimizes the number of tasks and

shortens the delay of tasks

Resourceallocation

A framework to administrate resources and amethod to predict and administrate resources [50]

Assists service providers to predict the amount of availableresources based on different types of service customers anddeals with the phenomenon that objects or devices withdraw

from resource utilization at any time

A bi-matching approach [51] Improves the performance of the system and obtains highercost efficiency

A framework for joint optimization [52] Optimizes resource allocation and improves networkperformance

A method to allocate computing andcommunication resources [53]

Allocates computing and communication resources transferscomputing jobs to remote cloud and nodes simplifies edge

nodesrsquo computing and saves computing energyA platform to allocate resources accessible to

client devices [54] Enables rapid response to computing resources and allocatesthe resources of devices near the host

A framework to manage resources [55] Considers the fluctuation of the customer abandonmentprobability and service types

Low latency A distributed algorithm [5] Solves the problem of wide range optimization and allocatesresources jointly

Complexity 9

Li et al [69] proposed communication framework in 5Gand studied the problem of allocating power andchannels So the signal data in the channel can beavailable and the total energy efficiency can be maximumFormula (9) [69] shows how to calculate the energy ef-ficiency of a system

U 1113944K

k11113944

M

i1EES

iK + 1113944K

K11113944

N

i1EEA

jk (9)

where EESiK and EEA

jk respectively denote the energy ef-ficiency of sensor S and actuator A on channels e sets ofsensors actuators and channels are respectively repre-sented as s S1 S2 SM1113864 1113865 A A1 A2 AN1113864 1113865 andC C1 C2 CK1113864 1113865

Liu et al [70] studied the problem of allocating resourcesefficiently on IoT that supplies wireless power In thismethod users are first grouped into accessible channelsAnd then power distribution of users grouped in the samechannel is studied to improve throughput of the networkis method can allocate finite resources to a large group ofusers Ejaz and Ibnkahla [71] proposed the resource allo-cation framework with several bands under cognitive 5GIoT In the highly dynamic environment of the IoTmultiband method can manage resources more flexiblyand reduce more energy consumption In addition areconstruction approach with several levels is proposed toallocate resources reasonably for applications with dif-ferent needs of QoS Colistra et al [72] proposed aprotocol which is distributed and optimal to allocateresources in heterogeneous IoT Because this protocol hasexcellent adaptability when changing topology of net-work it can distribute resources evenly among nodes Jianet al [73] proposed a multilevel allocating resources al-gorithm for IoT communication using advanced tech-nology e algorithm uses hierarchical structure and hasfast data processing rate and very low latency in a satu-rated or not saturated environment

Zheng and Liu [74] proposed a new algorithm to allocatebandwidth dynamically for controlling remote computers inthe IoT is method can reduce the error of signal re-construction under the same bandwidth and make thebandwidth allocation of IoT more reasonable Gai and Qiu[75] used reinforcement learning mechanisms to allocateresources to achieve high Quality of Experience ismethod can effectively solve the resource allocation prob-lems caused by the mismatch of service quality and complexservice providing condition in the IoT

Table 4 shows a summary of the above works eproblem in Table 4 represents the realization of dynamic andflexible allocation of resources e resources here canrepresent channels bandwidth and power

23 Scientific Workflow Execution Implementing scientificworkflows especially in heterogeneous environments canreduce resource waste and reduce energy costs Scientificworkflow can be obtained by reasonably allocating resourcesand dynamically deploying VMs

231 Cloud Computing Xu et al [76] proposed a resourceallocation method called EnRealan to solve the problem ofenergy consumption e dynamic deployment of VMs isgenerally adopted to execute scientific workflows Bousselmiet al [77] proposed a scheduling method based on energyperception for executing scientific workflows in cloudcomputing At first an algorithm of splitting workflows forenergy minimization is presented which can achieve a highparallelism without huge energy consumption en aheuristic algorithm used to optimize cat swarm is proposedfor the created partitions e algorithm can minimize thetotal consumption of energy and the execution time ofworkflows Sonia et al [78] proposed a workflow schedulingmethod with several objects and hybrid particle swarmoptimization algorithm In addition a method for dy-namically scaling voltage and frequency is proposed emethod canmake the processors work at any voltage level soas to minimize the energy consumption in the process ofworkflow scheduling Both Bousselmi et al [77] and Soniaet al [78] use scheduling method to achieve scientificworkflows and study the problem of energy consumptione difference is that Bousselmi et al [77] focuses on in-tensive computing tasks while Sonia et al [78] focuses onworkflow scheduling on heterogeneous computing systems

Cao [79] established a scheduling algorithm of scientificworkflows with an objective of energy savingis algorithmcan enable service providers to gain high profits and reduceusersrsquo overhead at the same time Li et al [80] proposed ascheduling algorithm based on cloud computing which canminimize cost of performing workflows within a specifiedtime In addition the rented VM was modified to save costfurther Khaleel and Zhu [81] proposed a scheduling algo-rithm and took scientific workflows as a model to make fulluse of cloud resources and save energy Shi et al [82]designed a flexible resource allocation and job schedulingmechanism to implement scientific workflows Because thismechanism can implement scientific workflows withinprescribed budgets and deadlines so it can work better thanother mechanisms

Table 5 shows a summary of the abovementioned workse problems in Table 5 are derived from the imple-mentation of the scientific workflow VM deployment refersto the rational allocation of VMs Workflow scheduling refersto reducing the scheduling energy and time In addition itrefers to the scheduling of the workflow on heterogeneoussystems Cost reduction refers to reducing the cost of workflowexecution Effective implementation is about scientific work-flow execution within a specified budget and time

24 Server Optimization Server optimization is also a goodway to save energy e goal of optimizing the server can beachieved by uninstalling unnecessary servers or consoli-dating servers as well as by reasonably scheduling tasksUnlike QoS optimization server optimization aims to op-timize the number of used servers improves the energyefficiency of servers and consolidates servers However QoSoptimization studies how tomake users get better experienceand meet their needs

10 Complexity

241 Cloud Computing Ge et al [83] proposed a game-theoretic method and transformed the problem of mini-mizing energy into a congestion game All mobile devices inthis method are participants in the game e methodchooses a server to unload the computation tasks to optimizeQoS levels and save energy which can optimize the systemand save energy Wang et al [84] proposed a MapReduce-based multitask scheduling algorithm to achieve the ob-jective of energy saving is model is a two-layer modelwhich considers the impact of server performance changeson energy consumption and the limitation of networkbandwidth In addition a local search operator is designedbased on which a two-layer genetic algorithm is proposede algorithm can schedule tens of thousands of tasks incloud and achieve large-scale optimization Yanggratoke

et al [85] proposed a general generic gossip protocol aimingat allocating resources in cloud environment An instanti-ation of this protocol was developed to enable server con-solidation to allocate resources to more servers to meetchanging load patterns

25 LoadBalancing Load balancing can help save energy bymanaging the number of servers and allocating resources

251 Cloud Computing Paya and Marinescu [86] intro-duced an operation model that balances cloud computingload and expands applications to save energy e principleof this model is to define an operating system e systemshould make servers run in the system as many as possible

Table 5 Work summary of scientific workflow execution in cloud computing

Problems Solutions Literatures Advantages

VM deployment A resource allocation method namedEnRealan [76] Performs scientific workflows based on energy perception

across cloud platforms

Workflowscheduling

A scheduling method based on energyperception [77]

Achieves a high parallelism without huge energyconsumption and minimizes the total consumption of

energy and execution time of workflows

A workflow scheduling method with severalobjects and hybrid particle swarm

optimization algorithm[78]

Makes the processors work at any voltage level minimizesthe energy consumption in the process of workflowscheduling and studies the scheduling problem of

workflows on heterogeneous systemsA scheduling algorithm based on various

applications [79] Enables service providers to gain high profits and reducesuser overhead at the same time

Cost reduction A scheduling algorithm based on energyperception [80 81] Minimizes the cost of performing workflows while

meeting the time constraintEffectiveimplementation

A flexible resource allocation and jobscheduling mechanism [82] Implements scientific workflows within prescribed

budgets and deadlines

Table 4 Work summary of resource allocation and management in IoT

Problems Solutions Literatures Advantages

Resourcemanagement

A distributed cloud network framework [66]Replaces the centralized architecture with a distributed cloud

architecture solves the defects of the centralized cloudarchitecture and brings people better experience

A MILP model [68] Assigns the requirement of servicesrsquo resource to heterogeneousnetwork equipment interface

A framework for communication used in5G [69]

Transforms the resource allocation problem into a power andchannel allocation problem minimizes the total energy

consumption and improves QoS levelsA low complexity channel allocation

algorithm [70] Improves throughput of the network and allocates finiteresources to a large group of users

A resource allocation framework withseveral bands under cognitive 5G IoT [71] Manages resources more flexibly and reduces energy

consumption than common single-band approach

A protocol which is distributed andoptimal to allocate resources [72]

Has excellent adaptability when changing topology of networkand dynamically manages resources in the heterogeneous IoT

environmentA multilevel allocating resourcesalgorithm for IoT communication [73] Has fast data processing rate and very low latency in both

saturated and nonsaturated environment

A new algorithm to allocate bandwidthdynamically [74]

Reduces the error of signal reconstruction under the samebandwidth and makes the bandwidth allocation of IoT more

reasonable

A reinforcement learning mechanism [75]Effectively solves the resource allocation problems caused by themismatch of service quality and complex service providing

condition in the IoT

Complexity 11

When no tasks are being performed the system should adjustservers to sleep thus energy consumption can be reducedJustafort et al [87] mainly studied the problem of workloaddistribution across cloud computing environment and pro-posed a method to solve the problem of the VM layout So thefootprint of carbon can be effectively reduced Panwar andMallick [88] proposed an algorithm to dynamically manage theload and effectively distribute the total incoming requestsbetween VMs rough efficient and uniform utilization ofresources this algorithm can achieve uniform distribution ofload between servers Yang et al [89] proposed a powermanagement mechanism to balance the load e system canmonitor VMs and dynamically allocate the resources Yanget al [90] proposed an optimization system to better allocateresources dynamically which can balance the load of VMsrunning on multiple physical machines Under this systemVMs can be migrated automatically to adjust high and lowloads without interrupting services Yang et al [89 90] manageVMs to achieve load balancing ey allocate resources dy-namically to migrate VMs which can balance workloads ondifferent physical machines e difference is that Yanget al[89] integrate a dynamic resource allocation approachwithOpenNebulaWhile Yang et al [90] focuse on avoiding serviceoutages during VM migration

Table 6 shows a summary of the abovementioned workse problems in Table 6 are from the load balancing problemServer management is about the control of the number ofservers running in the system Workload management is therational allocation of workload or tasks VMmanagement refersto configuringVMresources andmigratingVMs to adjust loads

252 Fog Computing Xu et al [91] proposed a methodcalled ldquoDRAMrdquo to dynamically allocate resources in fogcomputing environment which can avoid both too highand too low loads e method first analyzes the loadbalance of different kinds of computing nodes And then itdesigns a fog resource allocation method to achieve loadbalance which allocates resources statically and migratesservices dynamically Oueis et al [92] studied the loadbalance problem in fog computing A custom fog clus-tering algorithm is proposed to solve the problem In theproblem several users need to offload computations andall of their demands need to be handled by local com-puting cluster

253 IoT Wang et al [93] established architecture of theenergy saving targeted system in industrial IoT In additionin order to predict sleep intervals they developed a sleepscheduling and a wake protocol which provide a better wayfor energy saving

3 Conclusion

is paper did a comprehensive study of QoS optimi-zation and energy saving in cloud computing edgecomputing fog computing and IoT models We sum-marized five main problems and analyzed their solutionsproposed by existing works By conducting this surveywe aim to help readers have a deeper understanding onthe concepts of different computing models and thetechniques of QoS optimization and energy saving inthese models

e investigated papers focus on issues about ensuringQoS and reducing SLA violations and resource manage-ment In the case of QoS assurance and SLA violationsreduction the main solution of QoS assurance is efficientVM management is solution can meet customersrsquo re-quirements through reasonable scheduling and integrationof VMs Most of resource management techniques are re-alized by reasonable scheduling of resources which canreduce the waste of VMs servers and traffic

Disclosure

is manuscript is an extension of A Survey of QoS Opti-mization and Energy Saving in Cloud Edge and IoT in e9th EAI International Conference on Cloud Computing

Conflicts of Interest

e authors declare that there are no conflicts of interestregarding the publication of this paper

Acknowledgments

is work was supported by the National Natural ScienceFoundation of China (Grant no 61702274) Natural ScienceFoundation of Jiangsu Province (Grant no BK20170958)and PAPD

Table 6 Work summary of load balancing in cloud computing

Problems Solutions Literatures AdvantagesServermanagement

An operation model that can balance cloudcomputing load and expand application [86] Saves energy by managing the number of servers running in

the system

Workloadmanagement

A hybrid approach [87] Reduces the footprint of carbon and allocates workload acrosscloud computing

An algorithm to dynamically manage theload [88] Manages the load evens the load distribution between

servers and allocates tasks between VMs

VMmanagement

A green power administration mechanism [89] Monitors and jointly allocates VM resources

An optimization system [90]Migrates VMs to adjust high and low loads without

interrupting services and balances the load of VMs runningon multiple physical machines

12 Complexity

References

[1] Q Zhang L Cheng and R Boutaba ldquoCloud computingstate-of-the-art and research challengesrdquo Journal of InternetServices and Applications vol 1 no 1 pp 7ndash18 2010

[2] W Shi J Cao Q Zhang Y Li and L Xu ldquoEdge computingvision and challengesrdquo IEEE Internet of Aings Journal vol 3no 5 pp 637ndash646 2016

[3] M Iorga L Feldman R Barton M J Martin N S Gorenand C Mahmoudi ldquoFog computing conceptual modelrdquo TechRep Recommendations of the National Institute of Standardsand Technology Gaithersburg MD USA 2018

[4] Nebbiolo ldquoFog vs edge computingrdquo Tech Rep NebbioloTechnologies Inc Milpitas CA USA 2018

[5] C T Do N H Tran C Pham M G R Alam J H Son andC S Hong ldquoA proximal algorithm for joint resource allo-cation andminimizing carbon footprint in geo-distributed fogcomputingrdquo in Proceedings of the 2015 International Con-ference on Information Networking (ICOIN) pp 324ndash329IEEE Siem Reap Cambodia January 2015

[6] S Vashi J Ram J Modi S Verma and C Prakash ldquoInternetof things (IoT) a vision architectural elements and securityissuesrdquo in Proceedings of the 2017 International Conference onI-SMAC (IoT in Social Mobile Analytics and Cloud)(I-SMAC) pp 492ndash496 IEEE Coimbatore India February2017

[7] M Mazzucco D Dyachuk and R Deters ldquoMaximizing cloudprovidersrsquo revenues via energy aware allocation policiesrdquo inProceedings of the 2010 IEEE 3rd International Conference onCloud Computing pp 131ndash138 IEEE Miami FL USA July2010

[8] Q He R Zhou X Zhang et al ldquoKeyword search for buildingservice-based systemsrdquo IEEE Transactions on Software En-gineering vol 43 no 7 pp 658ndash674 2016

[9] L Sun H Dong O K Hussain F K Hussain and A X LiuldquoA framework of cloud service selection with criteria inter-actionsrdquo Future Generation Computer Systems vol 94pp 749ndash764 2019

[10] M Mazzucco and D Dyachuk ldquoOptimizing cloud providersrevenues via energy efficient server allocationrdquo SustainableComputing Informatics and Systems vol 2 no 1 pp 1ndash122012

[11] Q He J Han F Chen et al ldquoQos-aware service selection forcustomisable multi-tenant service-based systems maturityand approachesrdquo in Proceedings of the 2015 IEEE 8th Inter-national Conference on Cloud Computing pp 237ndash244 IEEENew York NY USA July 2015

[12] L Sun J Ma H Wang Y Zhang and J Yong ldquoCloud servicedescription model an extension of usdl for cloud servicesrdquoIEEE Transactions on Services Computing vol 11 no 2pp 354ndash368 2015

[13] Y Wang Q He D Ye and Y Yang ldquoFormulating criticality-based cost-effective fault tolerance strategies for multi-tenantservice-based systemsrdquo IEEE Transactions on Software En-gineering vol 44 no 3 pp 291ndash307 2017

[14] S Mustafa K Bilal S U R Malik and S A Madani ldquoSla-aware energy efficient resource management for cloud en-vironmentsrdquo IEEE Access vol 6 pp 15004ndash15020 2018

[15] J Bi H Yuan M Tie and W Tan ldquoSla-based optimisation ofvirtualised resource for multi-tier web applications in clouddata centresrdquo Enterprise Information Systems vol 9 no 7pp 743ndash767 2015

[16] S Singh I Chana and R Buyya ldquoStar sla-aware autonomicmanagement of cloud resourcesrdquo IEEE Transactions on CloudComputing p 1 2017

[17] A Beloglazov and R Buyya ldquoEnergy efficient resourcemanagement in virtualized cloud data centersrdquo in Proceedingsof the 2010 10th IEEEACM International Conference onCluster Cloud and Grid Computing pp 826ndash831 IEEEComputer Society Melbourne Australia May 2010

[18] M Guazzone C Anglano and M Canonico ldquoEnergy-effi-cient resource management for cloud computing infra-structuresrdquo in Proceedings of the 2011 IEEE AirdInternational Conference on Cloud Computing Technology andScience pp 424ndash431 IEEE Athens Greece November 2011

[19] Y Sun J White and S Eade ldquoA model-based system toautomate cloud resource allocation and optimizationrdquo inProceedings of the International Conference on Model DrivenEngineering Languages and Systems pp 18ndash34 SpringerValencia Spain October 2014

[20] G Siddesh and K Srinivasa ldquoSla-driven dynamic resourceallocation on cloudsrdquo in Proceedings of the InternationalConference on Advanced Computing Networking and Securitypp 9ndash18 Springer Surathkal India December 2011

[21] S K Garg S K Gopalaiyengar and R Buyya ldquoSla-basedresource provisioning for heterogeneous workloads in avirtualized cloud datacenterrdquo in Proceedings of the Interna-tional Conference on Algorithms and Architectures for ParallelProcessing pp 371ndash384 Springer Melbourne AustraliaOctober 2011

[22] J Bi Z Zhu and H Yuan ldquoSla-aware dynamic resourceprovisioning for profit maximization in shared cloud datacentersrdquo in Proceedings of the International Conference onHigh Performance Networking Computing and Communi-cation Systems pp 366ndash372 Springer Singapore May 2011

[23] L Qi Y Chen Y Yuan S Fu X Zhang and X Xu ldquoA Qos-aware virtual machine scheduling method for energy con-servation in cloud-based cyber-physical systemsrdquo WorldWide Web vol 4 no 3 pp 1ndash23 2019

[24] A Beloglazov J Abawajy and R Buyya ldquoEnergy-aware re-source allocation heuristics for efficient management of datacenters for cloud computingrdquo Future Generation ComputerSystems vol 28 no 5 pp 755ndash768 2012

[25] L Qi R Wang C Hu S Li Q He and X Xu ldquoTime-awaredistributed service recommendation with privacy-preserva-tionrdquo Information Sciences vol 480 pp 354ndash364 2019

[26] Z Zhai B Cheng Y Tian J Chen L Zhao and M Niu ldquoAdata-driven service creation approach for end-usersrdquo IEEEAccess vol 4 pp 9923ndash9940 2016

[27] L Gu D Zeng S Guo A Barnawi and Y Xiang ldquoCostefficient resource management in fog computing supportedmedical cyber-physical systemrdquo IEEE Transactions onEmerging Topics in Computing vol 5 no 1 pp 108ndash119 2015

[28] L Ni J Zhang C Jiang C Yan and K Yu ldquoResource al-location strategy in fog computing based on priced timed petrinetsrdquo Ieee Internet of Aings Journal vol 4 no 5pp 1216ndash1228 2017

[29] L Wei T Yang F C Delicato et al ldquoOn enabling sustainableedge computing with renewable energy resourcesrdquo IEEECommunications Magazine vol 56 no 5 pp 94ndash101 2018

[30] P Lai Q He G Cui et al ldquoEdge user allocation with dynamicquality of servicerdquo in Proceedings of the International Con-ference on Service-Oriented Computing pp 86ndash101 SpringerToulouse France October 2019

[31] X Xu Q Liu X Zhang J Zhang L Qi and W Dou ldquoAblockchain-powered crowdsourcing method with privacy

Complexity 13

preservation in mobile environmentrdquo IEEE Transactions onComputational Social Systems vol 6 no 6 pp 1407ndash1419 2019

[32] O Rolik E Zharikov and S Telenyk ldquoMicrocloud-basedarchitecture of management system for IoT infrastructuresrdquoin Proceedings of the 2016 Aird International Scientific-Practical Conference Problems of Infocommunications Scienceand Technology (PIC SampT) pp 149ndash151 IEEE KharkivUkaraine October 2016

[33] W He S Guo Y Liang R Ma X Qiu and L Shi ldquoQos-awareand resource-efficient dynamic slicingmechanism for internetof thingsrdquo Computers Materials amp Continua vol 61 no 3pp 1345ndash1364 2019

[34] J Yao and N Ansari ldquoQos-aware fog resource provisioningand mobile device power control in IoT networksrdquo IEEETransactions on Network and Service Management vol 16no 1 pp 167ndash175 2018

[35] W Wang Y Jiang and W Wu ldquoMultiagent-based resourceallocation for energy minimization in cloud computing sys-temsrdquo IEEE Transactions on Systems Man and CyberneticsSystems vol 47 no 2 pp 1ndash16 2016

[36] K Krishnajyothi ldquoParallel data processing for effective dy-namic resource allocation in the cloudrdquo International Journalof Computer Applications vol 70 no 22 pp 1ndash4 2013

[37] M M Hassan B Song M S Hossain and A Alamri ldquoEf-ficient resource scheduling for big data processing in cloudplatformrdquo in Proceedings of the International Conference onInternet and Distributed Computing Systems pp 51ndash63Springer Calabria Italy September 2014

[38] C-M Wu R-S Chang and H-Y Chan ldquoA green energy-efficient scheduling algorithm using the dvfs technique forcloud datacentersrdquo Future Generation Computer Systemsvol 37 no 7 pp 141ndash147 2014

[39] Z Wang and X Su ldquoDynamically hierarchical resource-al-location algorithm in cloud computing environmentrdquo AeJournal of Supercomputing vol 71 no 7 pp 2748ndash2766 2015

[40] W-Y Lin G-Y Lin and H-Y Wei ldquoDynamic auctionmechanism for cloud resource allocationrdquo in Proceedings ofthe 2010 10th IEEEACM International Conference on ClusterCloud and Grid Computing pp 591-592 IEEE ComputerSociety Victoria Australia May 2010

[41] Y O Yazir C Matthews R Farahbod et al ldquoDynamic re-source allocation in computing clouds using distributedmultiple criteria decision analysisrdquo in Proceedings of the 2010IEEE 3rd International Conference on Cloud Computingpp 91ndash98 IEEE Washington DC USA July 2010

[42] W Lin J Z Wang C Liang and D Qi ldquoA threshold-baseddynamic resource allocation scheme for cloud computingrdquoProcedia Engineering vol 23 no 5 pp 695ndash703 2011

[43] X Xu S Fu L Qi et al ldquoAn IoT-oriented data placementmethod with privacy preservation in cloud environmentrdquoJournal of Network and Computer Applications vol 124pp 148ndash157 2018

[44] J Bokyun P Md Jalil L Daeho and S Doug Young ldquoEf-ficient computation offloading in mobile cloud computing forvideo streaming over 5 grdquo Computers Materials amp Continuavol 61 no 2 pp 439ndash463 2019

[45] H Sarbazi-Azad and A Y Zomaya ldquoEnergy-efficient resourceutilization in cloud computingrdquo in Large Scale Network-Centric Distributed Systems pp 377ndash408 Wiley-IEEE PressHoboken NJ USA 2014

[46] C-H Hsu K D Slagter S-C Chen and Y-C ChungldquoOptimizing energy consumption with task consolidation incloudsrdquo Information Sciences vol 258 no 3 pp 452ndash4622014

[47] C-H Hsu S-C Chen C-C Lee et al ldquoEnergy-aware taskconsolidation technique for cloud computingrdquo in Proceedingsof the 2011 IEEE Aird International Conference on CloudComputing Technology and Science pp 115ndash121 IEEE Ath-ens Greece 2011

[48] S K Panda and P K Jana ldquoAn efficient task consolidationalgorithm for cloud computing systemsrdquo in Proceedings of theInternational Conference on Distributed Computing and In-ternet Technology pp 61ndash74 Springer Bhubaneswar IndiaJanuary 2016

[49] L Yin J Luo and H Luo ldquoTasks scheduling and resourceallocation in fog computing based on containers for smartmanufacturingrdquo IEEE Transactions on Industrial Informaticsvol 14 no 10 pp 4712ndash4721 2018

[50] M Aazam and E-N Huh ldquoDynamic resource provisioningthrough fog micro datacenterrdquo in Proceedings of the 2015IEEE International Conference on Pervasive Computing andCommunication Workshops (PerCom Workshops) pp 105ndash110 IEEE St Louis MO USA March 2015

[51] B Jia H Hu Y Zeng T Xu and Y Yang ldquoDouble-matchingresource allocation strategy in fog computing networks basedon cost efficiencyrdquo Journal of Communications and Networksvol 20 no 3 pp 237ndash246 2018

[52] H Zhang Y Xiao S Bu D Niyato F R Yu and Z HanldquoComputing resource allocation in three-tier IoT fog net-works a joint optimization approach combining stackelberggame and matchingrdquo IEEE Internet of Aings Journal vol 4no 5 pp 1204ndash1215 2017

[53] J Tan T-H Chang and T Q Quelc ldquoMinimum energyresource allocation in fog radio access network with fronthauland latency constraintsrdquo in Proceedings of the 2018 IEEE 19thInternational Workshop on Signal Processing Advances inWireless Communications (SPAWC) pp 1ndash5 IEEE KalamataGreece June 2018

[54] D R de Vasconcelos R M de Castro Andrade andJ N de Souza ldquoSmart shadowndashan autonomous availabilitycomputation resource allocation platform for internet ofthings in the fog computing environmentrdquo in Proceedings of the2015 International Conference on Distributed Computing inSensor Systems pp 216-217 IEEE Fortaleza Brazil June 2015

[55] M Aazam M St-Hilaire C-H Lung and I LambadarisldquoPre-fog IoT trace based probabilistic resource estimation atfogrdquo in Proceedings of the 2016 13th IEEE Annual ConsumerCommunications amp Networking Conference (CCNC) pp 12ndash17 IEEE Las Vegas NV USA January 2016

[56] N D Tung L L Bao and B Vijay ldquoPrice-based resourceallocation for edge computing a market equilibrium ap-proachrdquo IEEE Transactions on Cloud Computing p 1 2018

[57] X Xu C He Z Xu L Qi S Wan and M Z A BhuiyanldquoJoint optimization of offloading utility and privacy for edgecomputing enabled IoTrdquo IEEE Internet of Aings Journal2019

[58] X Xu Y Chen X Zhang Q Liu X Liu and L Qi ldquoAblockchain-based computation offloading method for edgecomputing in 5G networksrdquo Software Practice and Experi-ence 2019

[59] X Xu Y Li T Huang et al ldquoAn energy-aware computationoffloading method for smart edge computing in wirelessmetropolitan area networksrdquo Journal of Network and Com-puter Applications vol 133 pp 75ndash85 2019

[60] X Xu Y Xue L Qi et al ldquoAn edge computing-enabledcomputation offloading method with privacy preservation forinternet of connected vehiclesrdquo Future Generation ComputerSystems vol 96 pp 89ndash100 2019

14 Complexity

[61] G Yeting L Fang X Nong and C Zhengguo ldquoTask-basedresource allocation bid in edge computing micro datacenterrdquoComputers Materials amp Continua vol 61 no 2 pp 777ndash7922019

[62] X Chen L Jiao W Li and X Fu ldquoEfficient multi-usercomputation offloading for mobile-edge cloud computingrdquoIEEEACM Transactions on Networking vol 24 no 5pp 2795ndash2808 2016

[63] B Gao L He X Lu C Chang K Li and K Li ldquoDevelopingenergy-aware task allocation schemes in cloud-assisted mo-bile workflowsrdquo in Proceedings of the 2015 IEEE InternationalConference on Computer and Information Technology Ubiq-uitous Computing and Communications Dependable Auto-nomic and Secure Computing Pervasive Intelligence andComputing pp 1266ndash1273 IEEE Liverpool UK October2015

[64] X Xu X Zhang H Gao Y Xue L Qi andW Dou ldquoBecomeblockchain-enabled computation offloading for IoT in mobileedge computingrdquo IEEE Transactions on Industrial Infor-matics 2019

[65] W Yifei W Zhaoying G Da and Y F Richard ldquoDeepq-learning based computation offloading strategy for mobileedge computingrdquo Computers Materials amp Continua vol 59no 1 pp 89ndash104 2019

[66] M Barcelo A Correa J Llorca A M Tulino J L Vicarioand A Morell ldquoIoT-cloud service optimization in nextgeneration smart environmentsrdquo IEEE Journal on SelectedAreas in Communications vol 34 no 12 pp 4077ndash40902016

[67] B Cheng M Wang S Zhao Z Zhai D Zhu and J ChenldquoSituation-aware dynamic service coordination in an IoTenvironmentrdquo IEEEACM Transactions On Networkingvol 25 no 4 pp 2082ndash2095 2017

[68] V Angelakis I Avgouleas N Pappas and D Yuan ldquoFlexibleallocation of heterogeneous resources to services on an IoTdevicerdquo in Proceedings of the 2015 IEEE Conference onComputer Communications Workshops (INFOCOMWKSHPS) pp 99-100 IEEE Hong Kong China April 2015

[69] S Li Q Ni Y Sun G Min and S Al-Rubaye ldquoEnergy-ef-ficient resource allocation for industrial cyber-physical IoTsystems in 5g erardquo IEEE Transactions on Industrial Infor-matics vol 14 no 6 pp 2618ndash2628 2018

[70] X Liu Z Qin Y Gao and J A McCann ldquoResource allocationin wireless powered IoT networksrdquo IEEE Internet of AingsJournal vol 6 no 3 pp 4935ndash4945 2019

[71] W Ejaz and M Ibnkahla ldquoMulti-band spectrum sensing andresource allocation for IoT in cognitive 5g networksrdquo IEEEInternet of Aings Journal vol 5 no 1 pp 150ndash163 2017

[72] G Colistra V Pilloni and L Atzori ldquoTask allocation in groupof nodes in the IoT a consensus approachrdquo in Proceedings ofthe 2014 IEEE International Conference on Communications(ICC) pp 3848ndash3853 IEEE Sydney Australia June 2014

[73] J Li Q Sun and G Fan ldquoResource allocation for multiclassservice in IoTuplink communicationsrdquo inProceedings of the 20163rd International Conference on Systems and Informatics (ICSAI)pp 777ndash781 IEEE Shanghai China November 2016

[74] L I Zheng and K H Liu ldquoDynamic bandwidth resourceallocation algorithm in internet of things and its applicationrdquoComputer Engineering vol 38 no 17 pp 16ndash19 2012

[75] K Gai and M Qiu ldquoOptimal resource allocation using re-inforcement learning for IoT content-centric servicesrdquo Ap-plied Soft Computing vol 70 pp 12ndash21 2018

[76] X Xu W Dou X Zhang and J Chen ldquoEnreal an energy-aware resource allocation method for scientific workflow

executions in cloud environmentrdquo IEEE Transactions onCloud Computing vol 4 no 2 pp 166ndash179 2016

[77] K Bousselmi Z Brahmi and M M Gammoudi ldquoEnergyefficient partitioning and scheduling approach for scientificworkflows in the cloudrdquo in Proceedings of the 2016 IEEEInternational Conference on Services Computing (SCC)pp 146ndash154 IEEE San Francisco CA USA June 2016

[78] Y Sonia C Rachid K Hubert and G Bertrand ldquoMulti-objective approach for energy-aware workflow scheduling incloud computing environmentsrdquo Ae Scientific World Jour-nal vol 2013 Article ID 350934 13 pages 2013

[79] F Cao Efficient Scientific Workflow Scheduling in CloudEnvironment Southern Illinois University Carbondale ILUSA 2014

[80] Z Li J Ge H Hu W Song H Hu and B Luo ldquoCost andenergy aware scheduling algorithm for scientific workflowswith deadline constraint in cloudsrdquo IEEE Transactions onServices Computing vol 11 no 4 pp 713ndash726 2015

[81] M Khaleel and M M Zhu ldquoEnergy-aware job managementapproaches for workflow in cloudrdquo in Proceedings of the 2015IEEE International Conference on Cluster Computingpp 506-507 IEEE Chicago IL USA September 2015

[82] J Shi J Luo F Dong and J Zhang ldquoA budget and deadlineaware scientific workflow resource provisioning and sched-uling mechanism for cloudrdquo in Proceedings of the 2014 IEEE18th International Conference on Computer Supported Co-operative Work in Design (CSCWD) pp 672ndash677 IEEEHsinchu Taiwan January 2014

[83] Y Ge Y Zhang Q Qiu and Y-H Lu ldquoA game theoreticresource allocation for overall energy minimization in mobilecloud computing systemrdquo in Proceedings of the 2012ACMIEEE International Symposium on Low Power Electronicsand Design pp 279ndash284 ACM Redondo Beach CA USAAugust 2012

[84] X Wang Y Wang and C Yue ldquoAn energy-aware bi-leveloptimization model for multi-job scheduling problems undercloud computingrdquo Soft Computing vol 20 no 1 pp 303ndash3172016

[85] R Yanggratoke F Wuhib and R Stadler ldquoGossip-basedresource allocation for green computing in large cloudsrdquo inProceedings of the 2011 7th International Conference onNetwork and Service Management pp 1ndash9 IEEE ParisFrance October 2011

[86] A Paya and D C Marinescu ldquoEnergy-aware load balancingand application scaling for the cloud ecosystemrdquo IEEETransactions on Cloud Computing vol 5 no 1 pp 15ndash272015

[87] V D Justafort R Beaubrun and S Pierre ldquoA hybrid ap-proach for optimizing carbon footprint in intercloud envi-ronmentrdquo IEEE Transactions on Services Computing vol 12no 2 pp 186ndash198 2016

[88] R Panwar and B Mallick ldquoLoad balancing in cloud com-puting using dynamic load management algorithmrdquo inProceedings of the 2015 International Conference on GreenComputing and Internet of Aings (ICGCIoT) pp 773ndash778IEEE Noida India October 2015

[89] C-T Yang K-C Wang H-Y Cheng C-T Kuo andW C C Chu ldquoGreen power management with dynamicresource allocation for cloud virtual machinesrdquo in Proceedingsof the 2011 IEEE International Conference on High Perfor-mance Computing and Communications pp 726ndash733 IEEEAlberta Canada September 2011

[90] C-T Yang H-Y Cheng and K-L Huang ldquoA dynamicresource allocation model for virtual machine management

Complexity 15

on cloudrdquo in Proceedings of the International Conference onGrid and Distributed Computing pp 581ndash590 SpringerGangneug Korea December 2011

[91] X Xu F Shucun C Qing et al ldquoDynamic resource allocationfor load balancing in fog environmentrdquo Wireless Commu-nications amp Mobile Computing vol 2018 no 2 Article ID6421607 15 pages 2018

[92] J Oueis E C Strinati and S Barbarossa ldquoe fog balancingload distribution for small cell cloud computingrdquo in Pro-ceedings of the 2015 IEEE 81st Vehicular Technology Confer-ence (VTC Spring) pp 1ndash6 IEEE Glasgow UK May 2015

[93] K Wang Y Wang Y Sun S Guo and J Wu ldquoGreen in-dustrial internet of things architecture an energy-efficientperspectiverdquo IEEE Communications Magazine vol 54 no 12pp 48ndash54 2016

16 Complexity

merges tasks in virtual clusters Once a task migrationhappens the energy cost model will take into account thenetwork latency Sarbazi-Azad and Zomaya [45] and Hsuet al [46] both maximize the benefit of cloud resources byusing task merging techniques Sarbazi-Azad and Zomaya[45] uses a greedy algorithm called MaxUtil While Hsuet al [46] takes into account the network latency associatedwith task migration So in [46] a 17 improvement isachieved over MaxUtil

E0m Vi( 1113857 1113944m

t0Et Vi( 1113857 (6)

E0m VCk( 1113857 1113944n

i0E0m Vi( 1113857 (7)

where Et is the energy consumption in unit time and n is thenumber of VMs in the cluster

Hsu et al [47] proposed a task integration technologybased on the energy perception According to the charac-teristics of most cloud systems the principle of using 70CPU is proposed to administrate job integration amongvirtual clusters is technology is very effective in reducingthe amount of energy consumed in cloud systems bymerging tasks Panda and Jana [48] proposed an algorithmwith several criteria to combine tasks e algorithm notonly considers the time needed for processing jobs but alsoconsiders the utilization rate of VMs And the algorithm ismore energy efficient because it takes into account not onlythe processing time but also the utilization rate of VMsWang and Su [39] proposed a resource allocation algorithmto deal with wide range of communication between nodes incloud environment is algorithm uses recognition tech-nology to dynamically distribute jobs and nodes accordingto computing ability and factors of storage And it can re-duce the traffic when allocating resources because it usesdynamic hierarchy Lin et al [40] proposed a dynamicauction approach for resource allocation e approach canensure that even if there are many users and resourcesproviders will have reasonable profits and computing re-sources will be allocated correctly Yazir et al [41] proposeda new method to manage resources dynamically and au-tonomously Firstly resource management is split into jobsand each job is executed by autonomous nodes Secondautonomous nodes use the method called PROMETHEE toconfigure resources Krishnajyothi [36] proposed a frame-work which can implement parallel task processing to solvethe problem of low efficiency when submitting large tasksCompared with the static framework this framework candynamically allocate VMs thus reducing costs and the timeof processing tasks Lin et al [42] proposed a method toallocate resources dynamically by using thresholds Becausethis method uses the threshold value it can optimize thereallocation of resources improve the usage of resourcesand reduce the cost Xu et al [43] proposed a data placementstrategy named IDP for the data generated by IoTs devices toachieve reasonable data placement In this way the privacyof these data can be protected while resources are allocatedreasonably Jo et al [44] proposed a computing offload

framework under 5G network e framework transfers thecomputing burden to the cloud thus reducing the com-puting load of clients and the communication cost

Table 2 shows a summary of the abovementioned workse problem of resource allocation and management incloud computing can be divided into problems in Table 2VM management is about a reasonable configuration ofVMs Resource allocation represents the dynamic andflexible allocation of resources Task integration refers tocombining tasks to save energy and improve efficiency

222 Fog Computing Yin et al [49] established a newmodel of scheduling jobs which applies containers In orderto make sure that jobs can be finished on time a jobscheduling algorithm is developed e algorithm can alsooptimize the number of tasks that can be performed togetheron the nodes in fog computing And this paper proposes aredistribution mechanism to shorten the delay of tasksese methods are very effective in reducing task delaysAazam and Huh [50] established a framework to admin-istrate resources effectively in the mode of fog computingConsidering that there are various types of objects anddevices the connection between them may be volatile So amethod for predicting and administrating resources isproposed e method considers that any object or devicecan quit using resources at anytime Cuong et al [5] studiedthe allocating resources jointly problem and the problem ofcarbon footprint minimization in fog data center Formula(8) [5] is used to denote the energy consumption of serversIn addition a distributed algorithm is proposed to solve theproblem of wide range optimization

P(y) C middot Pidle + Ppeak minus Pidle1113872 1113873 middot y middot κ (8)

where P(y) represents the power supply required by theservers in a data center y represents the video stream κdenotes a conversion factor that converts the video streaminto workload C represents the data centerrsquos load capacityand Pidle and Ppeak respectively represent the idle power andpeak power of the servers

Jia et al [51] studied the problem of computing resourceallocation in fog computing network with three levelsFirstly the problem of resource allocation is transformedinto a bilateral matching optimal problem And then a bi-matching approach is proposed for this problem which canimprove the performance of the system and obtain highercost efficiency Zhang et al [52] proposed a framework forjoint optimization under fog computing to allocate fognodesrsquo finite computing resources e framework canachieve the best allocation and effectively improve thenetworksrsquo performance Tan et al [53] presented a method toallocate computing and communication resources emethod transfers computing jobs to remote cloud and nodesand simplifies edge nodesrsquo computing and computing en-ergy Vasconcelos et al [54] developed a platform to allocateresources accessible to client devices in fog computingenvironment allocating the resources of devices near thehost to meet the applications needs for rapid response tocomputing resources Aazam et al [55] presented a method

Complexity 7

to estimate and manage resource in fog computing emethod is based on the fluctuation of customer abandon-ment probability type and price of service and so on

Table 3 shows a summary of the abovementioned workse problems in Table 3 are also derived from resourceallocation and management problems Task allocationrepresents the scheduling and redistribution of tasks Re-source allocation is still about the dynamic and flexibleallocation of resources Low latency refers to taking shorttime to configure and manage resources which can improveefficiency

223 Edge Computing Tung et al [56] proposed a newframework for resource allocation based on market needse resources come from edge nodes (ENs) with limitedheterogeneous capabilities and are allocated to multiplecompeting services on the network edge Generating amarket equilibrium solution by reasonably pricing ENs canobtain the maximum utilization of marginal computingresources Xu et al [57] proposed a strategy to optimizeoffloading and privacy protection is strategy shifts tasksfirstly to improve the resource utilization of resource-limited

edge cells And then it balances QoS performance andprivacy protection to achieve joint optimization

Xu et al [58] proposed an offload strategy for edgecomputing under 5G network which uses block chaintechnologye optimal strategy is further obtained by usingthe balanced offloading method It solves the problem ofdata loss under the condition of transmission delay which iscaused by the uneven requirements of user equipments onresources Xu et al [59] proposed a computational off-loading method named EACO to reduce the energy con-sumption in smart computing models Figure 5 showsarchitecture of smart edge computing where the shortestpath is used to unload tasks EACO uses genetic algorithmsto reduce the energy consumption for operating edgecomputing nodes and improve the efficiency of performingcomplex computing tasks Xu et al [60] proposed a com-putational offloading strategy for edge computing to protectthe privacy of interconnected vehicle networks ey firstlyanalyzed privacy conflicts of tasks And then they designedthe communication route to obtain routing vehicles whichcan achieve the optimization of several objectives Yetinget al [61] proposed a unique resource allocation mechanisme mechanism takes each individual task as the basis for

Table 2 Work summary of resource allocation and management in cloud computing

Subproblems Solutions Literatures Advantages

VMmanagement

A VM allocation method based on severaldistributed agents [35] Centralizes VMs to physical machines and reduces the

overall energy costA framework which can implement parallel

task processing [36] Dynamically allocates VMs for large tasks

Resourceallocation

A general problem formula and a heuristicalgorithm with optimization parameters [37]

Dynamically allocates resources according to QoSrequirements and realizes energy saving by optimizing the

number of serversA scheduling algorithm based on thetechnology that can scale the voltage

frequency dynamically[38] Ensures the performance of executing jobs while

implementing green computing

A fuzzy pattern recognition technology [39] Reduces the traffic when allocating resources and usesdynamic hierarchy

A dynamic auction approach [40]Guarantees profits of providers and allocates resourcescorrectly when there are a large amount of users and

resources

PROMETHEE [41]Allows node agents to decompose and execute tasksautonomously to improve the flexibility of resource

allocation

A method using thresholds [42]Optimizes resource reallocation improves resource usagereduces cost and studies resource allocation strategies at the

application level

IoT-oriented data placement (IDP) method [43] Protects data privacy allocates resources reasonably andfocuses on the placement method of IoT data

A computing offload framework under 5Gnetwork [44] Reduces the computing load of clients and the

communication cost

Taskconsolidation

Two job consolidation heuristics namedldquoMaxUtilrdquo and energy-conscious task

consolidation[45] Promotes concurrent execution of multiple tasks and

improves energy efficiency

A job consolidation technique aiming atenergy saving [46 47]

Reduces the power consumption of cloud system protectsdata privacy allocates resources reasonably limits CPU

usage and merges tasks in virtual clusters

An algorithm with several criteria [48]Takes into account the job processing time and the VM

utilization rate simultaneously Dramatically reduces energyconsumption compared with state-of-the-art works

8 Complexity

resource allocation rather than for the whole service It reducesthe packet loss rate and saves energy by unloading services

224 MEC Chen et al [62] studied the problem of com-puting unloading with several users in the environment ofMEC with wireless interference which have many channels Inaddition a distributed algorithm for computing unloading isdevelopede algorithm can perform the unloading well evenwhen there are a large number of users Gao et al [63] built aquadratic binary program which is able to assign tasks inmobile cloud computing environment Two algorithms arepresented to obtain the optimal solution Both of these heuristicalgorithms can effectively solve the task assignment problemXu et al [64] proposed an offloadingmethod using block chaintechnology It can guarantee the loss of data in offloading tasksunder edge computing And it can solve the problem of re-source requests out of proportion due to the limited load of

edge computing equipment during task transfer Yifei et al [65]proposed a model-free reinforcement learning framework tosolve the problem of computational unloadingis model canbe applied to the computational unloading with time-changingcomputing requests

225 IoT Barcelo et al [66] expressed the problem ofservice allocation [67] as a mixed flow problem with min-imum cost which can be solved by LP solving this serviceallocation problem can solve the problems of unbalancednetwork load and delay of end-to-end service And it canalso figure out the problem of excessive consumption ofelectricity brought by the architecture of centralized cloudAngelakis et al [68] assigned the requirements of servicesresources to heterogeneous network interfaces of equipmentsSo more heterogeneous network interfaces can be used by alarge amount of services

Mobile devices

Mobile devices ECN

ECN

ECNECN

AP

AP

AP

APAP

Figure 5 An architecture of smart edge computing [59]

Table 3 Work summary of resource allocation and management in fog computing

Problems Solutions Literatures AdvantagesTaskallocation

A new model of scheduling jobs and aredistribution mechanism [49] Finishes jobs on time optimizes the number of tasks and

shortens the delay of tasks

Resourceallocation

A framework to administrate resources and amethod to predict and administrate resources [50]

Assists service providers to predict the amount of availableresources based on different types of service customers anddeals with the phenomenon that objects or devices withdraw

from resource utilization at any time

A bi-matching approach [51] Improves the performance of the system and obtains highercost efficiency

A framework for joint optimization [52] Optimizes resource allocation and improves networkperformance

A method to allocate computing andcommunication resources [53]

Allocates computing and communication resources transferscomputing jobs to remote cloud and nodes simplifies edge

nodesrsquo computing and saves computing energyA platform to allocate resources accessible to

client devices [54] Enables rapid response to computing resources and allocatesthe resources of devices near the host

A framework to manage resources [55] Considers the fluctuation of the customer abandonmentprobability and service types

Low latency A distributed algorithm [5] Solves the problem of wide range optimization and allocatesresources jointly

Complexity 9

Li et al [69] proposed communication framework in 5Gand studied the problem of allocating power andchannels So the signal data in the channel can beavailable and the total energy efficiency can be maximumFormula (9) [69] shows how to calculate the energy ef-ficiency of a system

U 1113944K

k11113944

M

i1EES

iK + 1113944K

K11113944

N

i1EEA

jk (9)

where EESiK and EEA

jk respectively denote the energy ef-ficiency of sensor S and actuator A on channels e sets ofsensors actuators and channels are respectively repre-sented as s S1 S2 SM1113864 1113865 A A1 A2 AN1113864 1113865 andC C1 C2 CK1113864 1113865

Liu et al [70] studied the problem of allocating resourcesefficiently on IoT that supplies wireless power In thismethod users are first grouped into accessible channelsAnd then power distribution of users grouped in the samechannel is studied to improve throughput of the networkis method can allocate finite resources to a large group ofusers Ejaz and Ibnkahla [71] proposed the resource allo-cation framework with several bands under cognitive 5GIoT In the highly dynamic environment of the IoTmultiband method can manage resources more flexiblyand reduce more energy consumption In addition areconstruction approach with several levels is proposed toallocate resources reasonably for applications with dif-ferent needs of QoS Colistra et al [72] proposed aprotocol which is distributed and optimal to allocateresources in heterogeneous IoT Because this protocol hasexcellent adaptability when changing topology of net-work it can distribute resources evenly among nodes Jianet al [73] proposed a multilevel allocating resources al-gorithm for IoT communication using advanced tech-nology e algorithm uses hierarchical structure and hasfast data processing rate and very low latency in a satu-rated or not saturated environment

Zheng and Liu [74] proposed a new algorithm to allocatebandwidth dynamically for controlling remote computers inthe IoT is method can reduce the error of signal re-construction under the same bandwidth and make thebandwidth allocation of IoT more reasonable Gai and Qiu[75] used reinforcement learning mechanisms to allocateresources to achieve high Quality of Experience ismethod can effectively solve the resource allocation prob-lems caused by the mismatch of service quality and complexservice providing condition in the IoT

Table 4 shows a summary of the above works eproblem in Table 4 represents the realization of dynamic andflexible allocation of resources e resources here canrepresent channels bandwidth and power

23 Scientific Workflow Execution Implementing scientificworkflows especially in heterogeneous environments canreduce resource waste and reduce energy costs Scientificworkflow can be obtained by reasonably allocating resourcesand dynamically deploying VMs

231 Cloud Computing Xu et al [76] proposed a resourceallocation method called EnRealan to solve the problem ofenergy consumption e dynamic deployment of VMs isgenerally adopted to execute scientific workflows Bousselmiet al [77] proposed a scheduling method based on energyperception for executing scientific workflows in cloudcomputing At first an algorithm of splitting workflows forenergy minimization is presented which can achieve a highparallelism without huge energy consumption en aheuristic algorithm used to optimize cat swarm is proposedfor the created partitions e algorithm can minimize thetotal consumption of energy and the execution time ofworkflows Sonia et al [78] proposed a workflow schedulingmethod with several objects and hybrid particle swarmoptimization algorithm In addition a method for dy-namically scaling voltage and frequency is proposed emethod canmake the processors work at any voltage level soas to minimize the energy consumption in the process ofworkflow scheduling Both Bousselmi et al [77] and Soniaet al [78] use scheduling method to achieve scientificworkflows and study the problem of energy consumptione difference is that Bousselmi et al [77] focuses on in-tensive computing tasks while Sonia et al [78] focuses onworkflow scheduling on heterogeneous computing systems

Cao [79] established a scheduling algorithm of scientificworkflows with an objective of energy savingis algorithmcan enable service providers to gain high profits and reduceusersrsquo overhead at the same time Li et al [80] proposed ascheduling algorithm based on cloud computing which canminimize cost of performing workflows within a specifiedtime In addition the rented VM was modified to save costfurther Khaleel and Zhu [81] proposed a scheduling algo-rithm and took scientific workflows as a model to make fulluse of cloud resources and save energy Shi et al [82]designed a flexible resource allocation and job schedulingmechanism to implement scientific workflows Because thismechanism can implement scientific workflows withinprescribed budgets and deadlines so it can work better thanother mechanisms

Table 5 shows a summary of the abovementioned workse problems in Table 5 are derived from the imple-mentation of the scientific workflow VM deployment refersto the rational allocation of VMs Workflow scheduling refersto reducing the scheduling energy and time In addition itrefers to the scheduling of the workflow on heterogeneoussystems Cost reduction refers to reducing the cost of workflowexecution Effective implementation is about scientific work-flow execution within a specified budget and time

24 Server Optimization Server optimization is also a goodway to save energy e goal of optimizing the server can beachieved by uninstalling unnecessary servers or consoli-dating servers as well as by reasonably scheduling tasksUnlike QoS optimization server optimization aims to op-timize the number of used servers improves the energyefficiency of servers and consolidates servers However QoSoptimization studies how tomake users get better experienceand meet their needs

10 Complexity

241 Cloud Computing Ge et al [83] proposed a game-theoretic method and transformed the problem of mini-mizing energy into a congestion game All mobile devices inthis method are participants in the game e methodchooses a server to unload the computation tasks to optimizeQoS levels and save energy which can optimize the systemand save energy Wang et al [84] proposed a MapReduce-based multitask scheduling algorithm to achieve the ob-jective of energy saving is model is a two-layer modelwhich considers the impact of server performance changeson energy consumption and the limitation of networkbandwidth In addition a local search operator is designedbased on which a two-layer genetic algorithm is proposede algorithm can schedule tens of thousands of tasks incloud and achieve large-scale optimization Yanggratoke

et al [85] proposed a general generic gossip protocol aimingat allocating resources in cloud environment An instanti-ation of this protocol was developed to enable server con-solidation to allocate resources to more servers to meetchanging load patterns

25 LoadBalancing Load balancing can help save energy bymanaging the number of servers and allocating resources

251 Cloud Computing Paya and Marinescu [86] intro-duced an operation model that balances cloud computingload and expands applications to save energy e principleof this model is to define an operating system e systemshould make servers run in the system as many as possible

Table 5 Work summary of scientific workflow execution in cloud computing

Problems Solutions Literatures Advantages

VM deployment A resource allocation method namedEnRealan [76] Performs scientific workflows based on energy perception

across cloud platforms

Workflowscheduling

A scheduling method based on energyperception [77]

Achieves a high parallelism without huge energyconsumption and minimizes the total consumption of

energy and execution time of workflows

A workflow scheduling method with severalobjects and hybrid particle swarm

optimization algorithm[78]

Makes the processors work at any voltage level minimizesthe energy consumption in the process of workflowscheduling and studies the scheduling problem of

workflows on heterogeneous systemsA scheduling algorithm based on various

applications [79] Enables service providers to gain high profits and reducesuser overhead at the same time

Cost reduction A scheduling algorithm based on energyperception [80 81] Minimizes the cost of performing workflows while

meeting the time constraintEffectiveimplementation

A flexible resource allocation and jobscheduling mechanism [82] Implements scientific workflows within prescribed

budgets and deadlines

Table 4 Work summary of resource allocation and management in IoT

Problems Solutions Literatures Advantages

Resourcemanagement

A distributed cloud network framework [66]Replaces the centralized architecture with a distributed cloud

architecture solves the defects of the centralized cloudarchitecture and brings people better experience

A MILP model [68] Assigns the requirement of servicesrsquo resource to heterogeneousnetwork equipment interface

A framework for communication used in5G [69]

Transforms the resource allocation problem into a power andchannel allocation problem minimizes the total energy

consumption and improves QoS levelsA low complexity channel allocation

algorithm [70] Improves throughput of the network and allocates finiteresources to a large group of users

A resource allocation framework withseveral bands under cognitive 5G IoT [71] Manages resources more flexibly and reduces energy

consumption than common single-band approach

A protocol which is distributed andoptimal to allocate resources [72]

Has excellent adaptability when changing topology of networkand dynamically manages resources in the heterogeneous IoT

environmentA multilevel allocating resourcesalgorithm for IoT communication [73] Has fast data processing rate and very low latency in both

saturated and nonsaturated environment

A new algorithm to allocate bandwidthdynamically [74]

Reduces the error of signal reconstruction under the samebandwidth and makes the bandwidth allocation of IoT more

reasonable

A reinforcement learning mechanism [75]Effectively solves the resource allocation problems caused by themismatch of service quality and complex service providing

condition in the IoT

Complexity 11

When no tasks are being performed the system should adjustservers to sleep thus energy consumption can be reducedJustafort et al [87] mainly studied the problem of workloaddistribution across cloud computing environment and pro-posed a method to solve the problem of the VM layout So thefootprint of carbon can be effectively reduced Panwar andMallick [88] proposed an algorithm to dynamically manage theload and effectively distribute the total incoming requestsbetween VMs rough efficient and uniform utilization ofresources this algorithm can achieve uniform distribution ofload between servers Yang et al [89] proposed a powermanagement mechanism to balance the load e system canmonitor VMs and dynamically allocate the resources Yanget al [90] proposed an optimization system to better allocateresources dynamically which can balance the load of VMsrunning on multiple physical machines Under this systemVMs can be migrated automatically to adjust high and lowloads without interrupting services Yang et al [89 90] manageVMs to achieve load balancing ey allocate resources dy-namically to migrate VMs which can balance workloads ondifferent physical machines e difference is that Yanget al[89] integrate a dynamic resource allocation approachwithOpenNebulaWhile Yang et al [90] focuse on avoiding serviceoutages during VM migration

Table 6 shows a summary of the abovementioned workse problems in Table 6 are from the load balancing problemServer management is about the control of the number ofservers running in the system Workload management is therational allocation of workload or tasks VMmanagement refersto configuringVMresources andmigratingVMs to adjust loads

252 Fog Computing Xu et al [91] proposed a methodcalled ldquoDRAMrdquo to dynamically allocate resources in fogcomputing environment which can avoid both too highand too low loads e method first analyzes the loadbalance of different kinds of computing nodes And then itdesigns a fog resource allocation method to achieve loadbalance which allocates resources statically and migratesservices dynamically Oueis et al [92] studied the loadbalance problem in fog computing A custom fog clus-tering algorithm is proposed to solve the problem In theproblem several users need to offload computations andall of their demands need to be handled by local com-puting cluster

253 IoT Wang et al [93] established architecture of theenergy saving targeted system in industrial IoT In additionin order to predict sleep intervals they developed a sleepscheduling and a wake protocol which provide a better wayfor energy saving

3 Conclusion

is paper did a comprehensive study of QoS optimi-zation and energy saving in cloud computing edgecomputing fog computing and IoT models We sum-marized five main problems and analyzed their solutionsproposed by existing works By conducting this surveywe aim to help readers have a deeper understanding onthe concepts of different computing models and thetechniques of QoS optimization and energy saving inthese models

e investigated papers focus on issues about ensuringQoS and reducing SLA violations and resource manage-ment In the case of QoS assurance and SLA violationsreduction the main solution of QoS assurance is efficientVM management is solution can meet customersrsquo re-quirements through reasonable scheduling and integrationof VMs Most of resource management techniques are re-alized by reasonable scheduling of resources which canreduce the waste of VMs servers and traffic

Disclosure

is manuscript is an extension of A Survey of QoS Opti-mization and Energy Saving in Cloud Edge and IoT in e9th EAI International Conference on Cloud Computing

Conflicts of Interest

e authors declare that there are no conflicts of interestregarding the publication of this paper

Acknowledgments

is work was supported by the National Natural ScienceFoundation of China (Grant no 61702274) Natural ScienceFoundation of Jiangsu Province (Grant no BK20170958)and PAPD

Table 6 Work summary of load balancing in cloud computing

Problems Solutions Literatures AdvantagesServermanagement

An operation model that can balance cloudcomputing load and expand application [86] Saves energy by managing the number of servers running in

the system

Workloadmanagement

A hybrid approach [87] Reduces the footprint of carbon and allocates workload acrosscloud computing

An algorithm to dynamically manage theload [88] Manages the load evens the load distribution between

servers and allocates tasks between VMs

VMmanagement

A green power administration mechanism [89] Monitors and jointly allocates VM resources

An optimization system [90]Migrates VMs to adjust high and low loads without

interrupting services and balances the load of VMs runningon multiple physical machines

12 Complexity

References

[1] Q Zhang L Cheng and R Boutaba ldquoCloud computingstate-of-the-art and research challengesrdquo Journal of InternetServices and Applications vol 1 no 1 pp 7ndash18 2010

[2] W Shi J Cao Q Zhang Y Li and L Xu ldquoEdge computingvision and challengesrdquo IEEE Internet of Aings Journal vol 3no 5 pp 637ndash646 2016

[3] M Iorga L Feldman R Barton M J Martin N S Gorenand C Mahmoudi ldquoFog computing conceptual modelrdquo TechRep Recommendations of the National Institute of Standardsand Technology Gaithersburg MD USA 2018

[4] Nebbiolo ldquoFog vs edge computingrdquo Tech Rep NebbioloTechnologies Inc Milpitas CA USA 2018

[5] C T Do N H Tran C Pham M G R Alam J H Son andC S Hong ldquoA proximal algorithm for joint resource allo-cation andminimizing carbon footprint in geo-distributed fogcomputingrdquo in Proceedings of the 2015 International Con-ference on Information Networking (ICOIN) pp 324ndash329IEEE Siem Reap Cambodia January 2015

[6] S Vashi J Ram J Modi S Verma and C Prakash ldquoInternetof things (IoT) a vision architectural elements and securityissuesrdquo in Proceedings of the 2017 International Conference onI-SMAC (IoT in Social Mobile Analytics and Cloud)(I-SMAC) pp 492ndash496 IEEE Coimbatore India February2017

[7] M Mazzucco D Dyachuk and R Deters ldquoMaximizing cloudprovidersrsquo revenues via energy aware allocation policiesrdquo inProceedings of the 2010 IEEE 3rd International Conference onCloud Computing pp 131ndash138 IEEE Miami FL USA July2010

[8] Q He R Zhou X Zhang et al ldquoKeyword search for buildingservice-based systemsrdquo IEEE Transactions on Software En-gineering vol 43 no 7 pp 658ndash674 2016

[9] L Sun H Dong O K Hussain F K Hussain and A X LiuldquoA framework of cloud service selection with criteria inter-actionsrdquo Future Generation Computer Systems vol 94pp 749ndash764 2019

[10] M Mazzucco and D Dyachuk ldquoOptimizing cloud providersrevenues via energy efficient server allocationrdquo SustainableComputing Informatics and Systems vol 2 no 1 pp 1ndash122012

[11] Q He J Han F Chen et al ldquoQos-aware service selection forcustomisable multi-tenant service-based systems maturityand approachesrdquo in Proceedings of the 2015 IEEE 8th Inter-national Conference on Cloud Computing pp 237ndash244 IEEENew York NY USA July 2015

[12] L Sun J Ma H Wang Y Zhang and J Yong ldquoCloud servicedescription model an extension of usdl for cloud servicesrdquoIEEE Transactions on Services Computing vol 11 no 2pp 354ndash368 2015

[13] Y Wang Q He D Ye and Y Yang ldquoFormulating criticality-based cost-effective fault tolerance strategies for multi-tenantservice-based systemsrdquo IEEE Transactions on Software En-gineering vol 44 no 3 pp 291ndash307 2017

[14] S Mustafa K Bilal S U R Malik and S A Madani ldquoSla-aware energy efficient resource management for cloud en-vironmentsrdquo IEEE Access vol 6 pp 15004ndash15020 2018

[15] J Bi H Yuan M Tie and W Tan ldquoSla-based optimisation ofvirtualised resource for multi-tier web applications in clouddata centresrdquo Enterprise Information Systems vol 9 no 7pp 743ndash767 2015

[16] S Singh I Chana and R Buyya ldquoStar sla-aware autonomicmanagement of cloud resourcesrdquo IEEE Transactions on CloudComputing p 1 2017

[17] A Beloglazov and R Buyya ldquoEnergy efficient resourcemanagement in virtualized cloud data centersrdquo in Proceedingsof the 2010 10th IEEEACM International Conference onCluster Cloud and Grid Computing pp 826ndash831 IEEEComputer Society Melbourne Australia May 2010

[18] M Guazzone C Anglano and M Canonico ldquoEnergy-effi-cient resource management for cloud computing infra-structuresrdquo in Proceedings of the 2011 IEEE AirdInternational Conference on Cloud Computing Technology andScience pp 424ndash431 IEEE Athens Greece November 2011

[19] Y Sun J White and S Eade ldquoA model-based system toautomate cloud resource allocation and optimizationrdquo inProceedings of the International Conference on Model DrivenEngineering Languages and Systems pp 18ndash34 SpringerValencia Spain October 2014

[20] G Siddesh and K Srinivasa ldquoSla-driven dynamic resourceallocation on cloudsrdquo in Proceedings of the InternationalConference on Advanced Computing Networking and Securitypp 9ndash18 Springer Surathkal India December 2011

[21] S K Garg S K Gopalaiyengar and R Buyya ldquoSla-basedresource provisioning for heterogeneous workloads in avirtualized cloud datacenterrdquo in Proceedings of the Interna-tional Conference on Algorithms and Architectures for ParallelProcessing pp 371ndash384 Springer Melbourne AustraliaOctober 2011

[22] J Bi Z Zhu and H Yuan ldquoSla-aware dynamic resourceprovisioning for profit maximization in shared cloud datacentersrdquo in Proceedings of the International Conference onHigh Performance Networking Computing and Communi-cation Systems pp 366ndash372 Springer Singapore May 2011

[23] L Qi Y Chen Y Yuan S Fu X Zhang and X Xu ldquoA Qos-aware virtual machine scheduling method for energy con-servation in cloud-based cyber-physical systemsrdquo WorldWide Web vol 4 no 3 pp 1ndash23 2019

[24] A Beloglazov J Abawajy and R Buyya ldquoEnergy-aware re-source allocation heuristics for efficient management of datacenters for cloud computingrdquo Future Generation ComputerSystems vol 28 no 5 pp 755ndash768 2012

[25] L Qi R Wang C Hu S Li Q He and X Xu ldquoTime-awaredistributed service recommendation with privacy-preserva-tionrdquo Information Sciences vol 480 pp 354ndash364 2019

[26] Z Zhai B Cheng Y Tian J Chen L Zhao and M Niu ldquoAdata-driven service creation approach for end-usersrdquo IEEEAccess vol 4 pp 9923ndash9940 2016

[27] L Gu D Zeng S Guo A Barnawi and Y Xiang ldquoCostefficient resource management in fog computing supportedmedical cyber-physical systemrdquo IEEE Transactions onEmerging Topics in Computing vol 5 no 1 pp 108ndash119 2015

[28] L Ni J Zhang C Jiang C Yan and K Yu ldquoResource al-location strategy in fog computing based on priced timed petrinetsrdquo Ieee Internet of Aings Journal vol 4 no 5pp 1216ndash1228 2017

[29] L Wei T Yang F C Delicato et al ldquoOn enabling sustainableedge computing with renewable energy resourcesrdquo IEEECommunications Magazine vol 56 no 5 pp 94ndash101 2018

[30] P Lai Q He G Cui et al ldquoEdge user allocation with dynamicquality of servicerdquo in Proceedings of the International Con-ference on Service-Oriented Computing pp 86ndash101 SpringerToulouse France October 2019

[31] X Xu Q Liu X Zhang J Zhang L Qi and W Dou ldquoAblockchain-powered crowdsourcing method with privacy

Complexity 13

preservation in mobile environmentrdquo IEEE Transactions onComputational Social Systems vol 6 no 6 pp 1407ndash1419 2019

[32] O Rolik E Zharikov and S Telenyk ldquoMicrocloud-basedarchitecture of management system for IoT infrastructuresrdquoin Proceedings of the 2016 Aird International Scientific-Practical Conference Problems of Infocommunications Scienceand Technology (PIC SampT) pp 149ndash151 IEEE KharkivUkaraine October 2016

[33] W He S Guo Y Liang R Ma X Qiu and L Shi ldquoQos-awareand resource-efficient dynamic slicingmechanism for internetof thingsrdquo Computers Materials amp Continua vol 61 no 3pp 1345ndash1364 2019

[34] J Yao and N Ansari ldquoQos-aware fog resource provisioningand mobile device power control in IoT networksrdquo IEEETransactions on Network and Service Management vol 16no 1 pp 167ndash175 2018

[35] W Wang Y Jiang and W Wu ldquoMultiagent-based resourceallocation for energy minimization in cloud computing sys-temsrdquo IEEE Transactions on Systems Man and CyberneticsSystems vol 47 no 2 pp 1ndash16 2016

[36] K Krishnajyothi ldquoParallel data processing for effective dy-namic resource allocation in the cloudrdquo International Journalof Computer Applications vol 70 no 22 pp 1ndash4 2013

[37] M M Hassan B Song M S Hossain and A Alamri ldquoEf-ficient resource scheduling for big data processing in cloudplatformrdquo in Proceedings of the International Conference onInternet and Distributed Computing Systems pp 51ndash63Springer Calabria Italy September 2014

[38] C-M Wu R-S Chang and H-Y Chan ldquoA green energy-efficient scheduling algorithm using the dvfs technique forcloud datacentersrdquo Future Generation Computer Systemsvol 37 no 7 pp 141ndash147 2014

[39] Z Wang and X Su ldquoDynamically hierarchical resource-al-location algorithm in cloud computing environmentrdquo AeJournal of Supercomputing vol 71 no 7 pp 2748ndash2766 2015

[40] W-Y Lin G-Y Lin and H-Y Wei ldquoDynamic auctionmechanism for cloud resource allocationrdquo in Proceedings ofthe 2010 10th IEEEACM International Conference on ClusterCloud and Grid Computing pp 591-592 IEEE ComputerSociety Victoria Australia May 2010

[41] Y O Yazir C Matthews R Farahbod et al ldquoDynamic re-source allocation in computing clouds using distributedmultiple criteria decision analysisrdquo in Proceedings of the 2010IEEE 3rd International Conference on Cloud Computingpp 91ndash98 IEEE Washington DC USA July 2010

[42] W Lin J Z Wang C Liang and D Qi ldquoA threshold-baseddynamic resource allocation scheme for cloud computingrdquoProcedia Engineering vol 23 no 5 pp 695ndash703 2011

[43] X Xu S Fu L Qi et al ldquoAn IoT-oriented data placementmethod with privacy preservation in cloud environmentrdquoJournal of Network and Computer Applications vol 124pp 148ndash157 2018

[44] J Bokyun P Md Jalil L Daeho and S Doug Young ldquoEf-ficient computation offloading in mobile cloud computing forvideo streaming over 5 grdquo Computers Materials amp Continuavol 61 no 2 pp 439ndash463 2019

[45] H Sarbazi-Azad and A Y Zomaya ldquoEnergy-efficient resourceutilization in cloud computingrdquo in Large Scale Network-Centric Distributed Systems pp 377ndash408 Wiley-IEEE PressHoboken NJ USA 2014

[46] C-H Hsu K D Slagter S-C Chen and Y-C ChungldquoOptimizing energy consumption with task consolidation incloudsrdquo Information Sciences vol 258 no 3 pp 452ndash4622014

[47] C-H Hsu S-C Chen C-C Lee et al ldquoEnergy-aware taskconsolidation technique for cloud computingrdquo in Proceedingsof the 2011 IEEE Aird International Conference on CloudComputing Technology and Science pp 115ndash121 IEEE Ath-ens Greece 2011

[48] S K Panda and P K Jana ldquoAn efficient task consolidationalgorithm for cloud computing systemsrdquo in Proceedings of theInternational Conference on Distributed Computing and In-ternet Technology pp 61ndash74 Springer Bhubaneswar IndiaJanuary 2016

[49] L Yin J Luo and H Luo ldquoTasks scheduling and resourceallocation in fog computing based on containers for smartmanufacturingrdquo IEEE Transactions on Industrial Informaticsvol 14 no 10 pp 4712ndash4721 2018

[50] M Aazam and E-N Huh ldquoDynamic resource provisioningthrough fog micro datacenterrdquo in Proceedings of the 2015IEEE International Conference on Pervasive Computing andCommunication Workshops (PerCom Workshops) pp 105ndash110 IEEE St Louis MO USA March 2015

[51] B Jia H Hu Y Zeng T Xu and Y Yang ldquoDouble-matchingresource allocation strategy in fog computing networks basedon cost efficiencyrdquo Journal of Communications and Networksvol 20 no 3 pp 237ndash246 2018

[52] H Zhang Y Xiao S Bu D Niyato F R Yu and Z HanldquoComputing resource allocation in three-tier IoT fog net-works a joint optimization approach combining stackelberggame and matchingrdquo IEEE Internet of Aings Journal vol 4no 5 pp 1204ndash1215 2017

[53] J Tan T-H Chang and T Q Quelc ldquoMinimum energyresource allocation in fog radio access network with fronthauland latency constraintsrdquo in Proceedings of the 2018 IEEE 19thInternational Workshop on Signal Processing Advances inWireless Communications (SPAWC) pp 1ndash5 IEEE KalamataGreece June 2018

[54] D R de Vasconcelos R M de Castro Andrade andJ N de Souza ldquoSmart shadowndashan autonomous availabilitycomputation resource allocation platform for internet ofthings in the fog computing environmentrdquo in Proceedings of the2015 International Conference on Distributed Computing inSensor Systems pp 216-217 IEEE Fortaleza Brazil June 2015

[55] M Aazam M St-Hilaire C-H Lung and I LambadarisldquoPre-fog IoT trace based probabilistic resource estimation atfogrdquo in Proceedings of the 2016 13th IEEE Annual ConsumerCommunications amp Networking Conference (CCNC) pp 12ndash17 IEEE Las Vegas NV USA January 2016

[56] N D Tung L L Bao and B Vijay ldquoPrice-based resourceallocation for edge computing a market equilibrium ap-proachrdquo IEEE Transactions on Cloud Computing p 1 2018

[57] X Xu C He Z Xu L Qi S Wan and M Z A BhuiyanldquoJoint optimization of offloading utility and privacy for edgecomputing enabled IoTrdquo IEEE Internet of Aings Journal2019

[58] X Xu Y Chen X Zhang Q Liu X Liu and L Qi ldquoAblockchain-based computation offloading method for edgecomputing in 5G networksrdquo Software Practice and Experi-ence 2019

[59] X Xu Y Li T Huang et al ldquoAn energy-aware computationoffloading method for smart edge computing in wirelessmetropolitan area networksrdquo Journal of Network and Com-puter Applications vol 133 pp 75ndash85 2019

[60] X Xu Y Xue L Qi et al ldquoAn edge computing-enabledcomputation offloading method with privacy preservation forinternet of connected vehiclesrdquo Future Generation ComputerSystems vol 96 pp 89ndash100 2019

14 Complexity

[61] G Yeting L Fang X Nong and C Zhengguo ldquoTask-basedresource allocation bid in edge computing micro datacenterrdquoComputers Materials amp Continua vol 61 no 2 pp 777ndash7922019

[62] X Chen L Jiao W Li and X Fu ldquoEfficient multi-usercomputation offloading for mobile-edge cloud computingrdquoIEEEACM Transactions on Networking vol 24 no 5pp 2795ndash2808 2016

[63] B Gao L He X Lu C Chang K Li and K Li ldquoDevelopingenergy-aware task allocation schemes in cloud-assisted mo-bile workflowsrdquo in Proceedings of the 2015 IEEE InternationalConference on Computer and Information Technology Ubiq-uitous Computing and Communications Dependable Auto-nomic and Secure Computing Pervasive Intelligence andComputing pp 1266ndash1273 IEEE Liverpool UK October2015

[64] X Xu X Zhang H Gao Y Xue L Qi andW Dou ldquoBecomeblockchain-enabled computation offloading for IoT in mobileedge computingrdquo IEEE Transactions on Industrial Infor-matics 2019

[65] W Yifei W Zhaoying G Da and Y F Richard ldquoDeepq-learning based computation offloading strategy for mobileedge computingrdquo Computers Materials amp Continua vol 59no 1 pp 89ndash104 2019

[66] M Barcelo A Correa J Llorca A M Tulino J L Vicarioand A Morell ldquoIoT-cloud service optimization in nextgeneration smart environmentsrdquo IEEE Journal on SelectedAreas in Communications vol 34 no 12 pp 4077ndash40902016

[67] B Cheng M Wang S Zhao Z Zhai D Zhu and J ChenldquoSituation-aware dynamic service coordination in an IoTenvironmentrdquo IEEEACM Transactions On Networkingvol 25 no 4 pp 2082ndash2095 2017

[68] V Angelakis I Avgouleas N Pappas and D Yuan ldquoFlexibleallocation of heterogeneous resources to services on an IoTdevicerdquo in Proceedings of the 2015 IEEE Conference onComputer Communications Workshops (INFOCOMWKSHPS) pp 99-100 IEEE Hong Kong China April 2015

[69] S Li Q Ni Y Sun G Min and S Al-Rubaye ldquoEnergy-ef-ficient resource allocation for industrial cyber-physical IoTsystems in 5g erardquo IEEE Transactions on Industrial Infor-matics vol 14 no 6 pp 2618ndash2628 2018

[70] X Liu Z Qin Y Gao and J A McCann ldquoResource allocationin wireless powered IoT networksrdquo IEEE Internet of AingsJournal vol 6 no 3 pp 4935ndash4945 2019

[71] W Ejaz and M Ibnkahla ldquoMulti-band spectrum sensing andresource allocation for IoT in cognitive 5g networksrdquo IEEEInternet of Aings Journal vol 5 no 1 pp 150ndash163 2017

[72] G Colistra V Pilloni and L Atzori ldquoTask allocation in groupof nodes in the IoT a consensus approachrdquo in Proceedings ofthe 2014 IEEE International Conference on Communications(ICC) pp 3848ndash3853 IEEE Sydney Australia June 2014

[73] J Li Q Sun and G Fan ldquoResource allocation for multiclassservice in IoTuplink communicationsrdquo inProceedings of the 20163rd International Conference on Systems and Informatics (ICSAI)pp 777ndash781 IEEE Shanghai China November 2016

[74] L I Zheng and K H Liu ldquoDynamic bandwidth resourceallocation algorithm in internet of things and its applicationrdquoComputer Engineering vol 38 no 17 pp 16ndash19 2012

[75] K Gai and M Qiu ldquoOptimal resource allocation using re-inforcement learning for IoT content-centric servicesrdquo Ap-plied Soft Computing vol 70 pp 12ndash21 2018

[76] X Xu W Dou X Zhang and J Chen ldquoEnreal an energy-aware resource allocation method for scientific workflow

executions in cloud environmentrdquo IEEE Transactions onCloud Computing vol 4 no 2 pp 166ndash179 2016

[77] K Bousselmi Z Brahmi and M M Gammoudi ldquoEnergyefficient partitioning and scheduling approach for scientificworkflows in the cloudrdquo in Proceedings of the 2016 IEEEInternational Conference on Services Computing (SCC)pp 146ndash154 IEEE San Francisco CA USA June 2016

[78] Y Sonia C Rachid K Hubert and G Bertrand ldquoMulti-objective approach for energy-aware workflow scheduling incloud computing environmentsrdquo Ae Scientific World Jour-nal vol 2013 Article ID 350934 13 pages 2013

[79] F Cao Efficient Scientific Workflow Scheduling in CloudEnvironment Southern Illinois University Carbondale ILUSA 2014

[80] Z Li J Ge H Hu W Song H Hu and B Luo ldquoCost andenergy aware scheduling algorithm for scientific workflowswith deadline constraint in cloudsrdquo IEEE Transactions onServices Computing vol 11 no 4 pp 713ndash726 2015

[81] M Khaleel and M M Zhu ldquoEnergy-aware job managementapproaches for workflow in cloudrdquo in Proceedings of the 2015IEEE International Conference on Cluster Computingpp 506-507 IEEE Chicago IL USA September 2015

[82] J Shi J Luo F Dong and J Zhang ldquoA budget and deadlineaware scientific workflow resource provisioning and sched-uling mechanism for cloudrdquo in Proceedings of the 2014 IEEE18th International Conference on Computer Supported Co-operative Work in Design (CSCWD) pp 672ndash677 IEEEHsinchu Taiwan January 2014

[83] Y Ge Y Zhang Q Qiu and Y-H Lu ldquoA game theoreticresource allocation for overall energy minimization in mobilecloud computing systemrdquo in Proceedings of the 2012ACMIEEE International Symposium on Low Power Electronicsand Design pp 279ndash284 ACM Redondo Beach CA USAAugust 2012

[84] X Wang Y Wang and C Yue ldquoAn energy-aware bi-leveloptimization model for multi-job scheduling problems undercloud computingrdquo Soft Computing vol 20 no 1 pp 303ndash3172016

[85] R Yanggratoke F Wuhib and R Stadler ldquoGossip-basedresource allocation for green computing in large cloudsrdquo inProceedings of the 2011 7th International Conference onNetwork and Service Management pp 1ndash9 IEEE ParisFrance October 2011

[86] A Paya and D C Marinescu ldquoEnergy-aware load balancingand application scaling for the cloud ecosystemrdquo IEEETransactions on Cloud Computing vol 5 no 1 pp 15ndash272015

[87] V D Justafort R Beaubrun and S Pierre ldquoA hybrid ap-proach for optimizing carbon footprint in intercloud envi-ronmentrdquo IEEE Transactions on Services Computing vol 12no 2 pp 186ndash198 2016

[88] R Panwar and B Mallick ldquoLoad balancing in cloud com-puting using dynamic load management algorithmrdquo inProceedings of the 2015 International Conference on GreenComputing and Internet of Aings (ICGCIoT) pp 773ndash778IEEE Noida India October 2015

[89] C-T Yang K-C Wang H-Y Cheng C-T Kuo andW C C Chu ldquoGreen power management with dynamicresource allocation for cloud virtual machinesrdquo in Proceedingsof the 2011 IEEE International Conference on High Perfor-mance Computing and Communications pp 726ndash733 IEEEAlberta Canada September 2011

[90] C-T Yang H-Y Cheng and K-L Huang ldquoA dynamicresource allocation model for virtual machine management

Complexity 15

on cloudrdquo in Proceedings of the International Conference onGrid and Distributed Computing pp 581ndash590 SpringerGangneug Korea December 2011

[91] X Xu F Shucun C Qing et al ldquoDynamic resource allocationfor load balancing in fog environmentrdquo Wireless Commu-nications amp Mobile Computing vol 2018 no 2 Article ID6421607 15 pages 2018

[92] J Oueis E C Strinati and S Barbarossa ldquoe fog balancingload distribution for small cell cloud computingrdquo in Pro-ceedings of the 2015 IEEE 81st Vehicular Technology Confer-ence (VTC Spring) pp 1ndash6 IEEE Glasgow UK May 2015

[93] K Wang Y Wang Y Sun S Guo and J Wu ldquoGreen in-dustrial internet of things architecture an energy-efficientperspectiverdquo IEEE Communications Magazine vol 54 no 12pp 48ndash54 2016

16 Complexity

to estimate and manage resource in fog computing emethod is based on the fluctuation of customer abandon-ment probability type and price of service and so on

Table 3 shows a summary of the abovementioned workse problems in Table 3 are also derived from resourceallocation and management problems Task allocationrepresents the scheduling and redistribution of tasks Re-source allocation is still about the dynamic and flexibleallocation of resources Low latency refers to taking shorttime to configure and manage resources which can improveefficiency

223 Edge Computing Tung et al [56] proposed a newframework for resource allocation based on market needse resources come from edge nodes (ENs) with limitedheterogeneous capabilities and are allocated to multiplecompeting services on the network edge Generating amarket equilibrium solution by reasonably pricing ENs canobtain the maximum utilization of marginal computingresources Xu et al [57] proposed a strategy to optimizeoffloading and privacy protection is strategy shifts tasksfirstly to improve the resource utilization of resource-limited

edge cells And then it balances QoS performance andprivacy protection to achieve joint optimization

Xu et al [58] proposed an offload strategy for edgecomputing under 5G network which uses block chaintechnologye optimal strategy is further obtained by usingthe balanced offloading method It solves the problem ofdata loss under the condition of transmission delay which iscaused by the uneven requirements of user equipments onresources Xu et al [59] proposed a computational off-loading method named EACO to reduce the energy con-sumption in smart computing models Figure 5 showsarchitecture of smart edge computing where the shortestpath is used to unload tasks EACO uses genetic algorithmsto reduce the energy consumption for operating edgecomputing nodes and improve the efficiency of performingcomplex computing tasks Xu et al [60] proposed a com-putational offloading strategy for edge computing to protectthe privacy of interconnected vehicle networks ey firstlyanalyzed privacy conflicts of tasks And then they designedthe communication route to obtain routing vehicles whichcan achieve the optimization of several objectives Yetinget al [61] proposed a unique resource allocation mechanisme mechanism takes each individual task as the basis for

Table 2 Work summary of resource allocation and management in cloud computing

Subproblems Solutions Literatures Advantages

VMmanagement

A VM allocation method based on severaldistributed agents [35] Centralizes VMs to physical machines and reduces the

overall energy costA framework which can implement parallel

task processing [36] Dynamically allocates VMs for large tasks

Resourceallocation

A general problem formula and a heuristicalgorithm with optimization parameters [37]

Dynamically allocates resources according to QoSrequirements and realizes energy saving by optimizing the

number of serversA scheduling algorithm based on thetechnology that can scale the voltage

frequency dynamically[38] Ensures the performance of executing jobs while

implementing green computing

A fuzzy pattern recognition technology [39] Reduces the traffic when allocating resources and usesdynamic hierarchy

A dynamic auction approach [40]Guarantees profits of providers and allocates resourcescorrectly when there are a large amount of users and

resources

PROMETHEE [41]Allows node agents to decompose and execute tasksautonomously to improve the flexibility of resource

allocation

A method using thresholds [42]Optimizes resource reallocation improves resource usagereduces cost and studies resource allocation strategies at the

application level

IoT-oriented data placement (IDP) method [43] Protects data privacy allocates resources reasonably andfocuses on the placement method of IoT data

A computing offload framework under 5Gnetwork [44] Reduces the computing load of clients and the

communication cost

Taskconsolidation

Two job consolidation heuristics namedldquoMaxUtilrdquo and energy-conscious task

consolidation[45] Promotes concurrent execution of multiple tasks and

improves energy efficiency

A job consolidation technique aiming atenergy saving [46 47]

Reduces the power consumption of cloud system protectsdata privacy allocates resources reasonably limits CPU

usage and merges tasks in virtual clusters

An algorithm with several criteria [48]Takes into account the job processing time and the VM

utilization rate simultaneously Dramatically reduces energyconsumption compared with state-of-the-art works

8 Complexity

resource allocation rather than for the whole service It reducesthe packet loss rate and saves energy by unloading services

224 MEC Chen et al [62] studied the problem of com-puting unloading with several users in the environment ofMEC with wireless interference which have many channels Inaddition a distributed algorithm for computing unloading isdevelopede algorithm can perform the unloading well evenwhen there are a large number of users Gao et al [63] built aquadratic binary program which is able to assign tasks inmobile cloud computing environment Two algorithms arepresented to obtain the optimal solution Both of these heuristicalgorithms can effectively solve the task assignment problemXu et al [64] proposed an offloadingmethod using block chaintechnology It can guarantee the loss of data in offloading tasksunder edge computing And it can solve the problem of re-source requests out of proportion due to the limited load of

edge computing equipment during task transfer Yifei et al [65]proposed a model-free reinforcement learning framework tosolve the problem of computational unloadingis model canbe applied to the computational unloading with time-changingcomputing requests

225 IoT Barcelo et al [66] expressed the problem ofservice allocation [67] as a mixed flow problem with min-imum cost which can be solved by LP solving this serviceallocation problem can solve the problems of unbalancednetwork load and delay of end-to-end service And it canalso figure out the problem of excessive consumption ofelectricity brought by the architecture of centralized cloudAngelakis et al [68] assigned the requirements of servicesresources to heterogeneous network interfaces of equipmentsSo more heterogeneous network interfaces can be used by alarge amount of services

Mobile devices

Mobile devices ECN

ECN

ECNECN

AP

AP

AP

APAP

Figure 5 An architecture of smart edge computing [59]

Table 3 Work summary of resource allocation and management in fog computing

Problems Solutions Literatures AdvantagesTaskallocation

A new model of scheduling jobs and aredistribution mechanism [49] Finishes jobs on time optimizes the number of tasks and

shortens the delay of tasks

Resourceallocation

A framework to administrate resources and amethod to predict and administrate resources [50]

Assists service providers to predict the amount of availableresources based on different types of service customers anddeals with the phenomenon that objects or devices withdraw

from resource utilization at any time

A bi-matching approach [51] Improves the performance of the system and obtains highercost efficiency

A framework for joint optimization [52] Optimizes resource allocation and improves networkperformance

A method to allocate computing andcommunication resources [53]

Allocates computing and communication resources transferscomputing jobs to remote cloud and nodes simplifies edge

nodesrsquo computing and saves computing energyA platform to allocate resources accessible to

client devices [54] Enables rapid response to computing resources and allocatesthe resources of devices near the host

A framework to manage resources [55] Considers the fluctuation of the customer abandonmentprobability and service types

Low latency A distributed algorithm [5] Solves the problem of wide range optimization and allocatesresources jointly

Complexity 9

Li et al [69] proposed communication framework in 5Gand studied the problem of allocating power andchannels So the signal data in the channel can beavailable and the total energy efficiency can be maximumFormula (9) [69] shows how to calculate the energy ef-ficiency of a system

U 1113944K

k11113944

M

i1EES

iK + 1113944K

K11113944

N

i1EEA

jk (9)

where EESiK and EEA

jk respectively denote the energy ef-ficiency of sensor S and actuator A on channels e sets ofsensors actuators and channels are respectively repre-sented as s S1 S2 SM1113864 1113865 A A1 A2 AN1113864 1113865 andC C1 C2 CK1113864 1113865

Liu et al [70] studied the problem of allocating resourcesefficiently on IoT that supplies wireless power In thismethod users are first grouped into accessible channelsAnd then power distribution of users grouped in the samechannel is studied to improve throughput of the networkis method can allocate finite resources to a large group ofusers Ejaz and Ibnkahla [71] proposed the resource allo-cation framework with several bands under cognitive 5GIoT In the highly dynamic environment of the IoTmultiband method can manage resources more flexiblyand reduce more energy consumption In addition areconstruction approach with several levels is proposed toallocate resources reasonably for applications with dif-ferent needs of QoS Colistra et al [72] proposed aprotocol which is distributed and optimal to allocateresources in heterogeneous IoT Because this protocol hasexcellent adaptability when changing topology of net-work it can distribute resources evenly among nodes Jianet al [73] proposed a multilevel allocating resources al-gorithm for IoT communication using advanced tech-nology e algorithm uses hierarchical structure and hasfast data processing rate and very low latency in a satu-rated or not saturated environment

Zheng and Liu [74] proposed a new algorithm to allocatebandwidth dynamically for controlling remote computers inthe IoT is method can reduce the error of signal re-construction under the same bandwidth and make thebandwidth allocation of IoT more reasonable Gai and Qiu[75] used reinforcement learning mechanisms to allocateresources to achieve high Quality of Experience ismethod can effectively solve the resource allocation prob-lems caused by the mismatch of service quality and complexservice providing condition in the IoT

Table 4 shows a summary of the above works eproblem in Table 4 represents the realization of dynamic andflexible allocation of resources e resources here canrepresent channels bandwidth and power

23 Scientific Workflow Execution Implementing scientificworkflows especially in heterogeneous environments canreduce resource waste and reduce energy costs Scientificworkflow can be obtained by reasonably allocating resourcesand dynamically deploying VMs

231 Cloud Computing Xu et al [76] proposed a resourceallocation method called EnRealan to solve the problem ofenergy consumption e dynamic deployment of VMs isgenerally adopted to execute scientific workflows Bousselmiet al [77] proposed a scheduling method based on energyperception for executing scientific workflows in cloudcomputing At first an algorithm of splitting workflows forenergy minimization is presented which can achieve a highparallelism without huge energy consumption en aheuristic algorithm used to optimize cat swarm is proposedfor the created partitions e algorithm can minimize thetotal consumption of energy and the execution time ofworkflows Sonia et al [78] proposed a workflow schedulingmethod with several objects and hybrid particle swarmoptimization algorithm In addition a method for dy-namically scaling voltage and frequency is proposed emethod canmake the processors work at any voltage level soas to minimize the energy consumption in the process ofworkflow scheduling Both Bousselmi et al [77] and Soniaet al [78] use scheduling method to achieve scientificworkflows and study the problem of energy consumptione difference is that Bousselmi et al [77] focuses on in-tensive computing tasks while Sonia et al [78] focuses onworkflow scheduling on heterogeneous computing systems

Cao [79] established a scheduling algorithm of scientificworkflows with an objective of energy savingis algorithmcan enable service providers to gain high profits and reduceusersrsquo overhead at the same time Li et al [80] proposed ascheduling algorithm based on cloud computing which canminimize cost of performing workflows within a specifiedtime In addition the rented VM was modified to save costfurther Khaleel and Zhu [81] proposed a scheduling algo-rithm and took scientific workflows as a model to make fulluse of cloud resources and save energy Shi et al [82]designed a flexible resource allocation and job schedulingmechanism to implement scientific workflows Because thismechanism can implement scientific workflows withinprescribed budgets and deadlines so it can work better thanother mechanisms

Table 5 shows a summary of the abovementioned workse problems in Table 5 are derived from the imple-mentation of the scientific workflow VM deployment refersto the rational allocation of VMs Workflow scheduling refersto reducing the scheduling energy and time In addition itrefers to the scheduling of the workflow on heterogeneoussystems Cost reduction refers to reducing the cost of workflowexecution Effective implementation is about scientific work-flow execution within a specified budget and time

24 Server Optimization Server optimization is also a goodway to save energy e goal of optimizing the server can beachieved by uninstalling unnecessary servers or consoli-dating servers as well as by reasonably scheduling tasksUnlike QoS optimization server optimization aims to op-timize the number of used servers improves the energyefficiency of servers and consolidates servers However QoSoptimization studies how tomake users get better experienceand meet their needs

10 Complexity

241 Cloud Computing Ge et al [83] proposed a game-theoretic method and transformed the problem of mini-mizing energy into a congestion game All mobile devices inthis method are participants in the game e methodchooses a server to unload the computation tasks to optimizeQoS levels and save energy which can optimize the systemand save energy Wang et al [84] proposed a MapReduce-based multitask scheduling algorithm to achieve the ob-jective of energy saving is model is a two-layer modelwhich considers the impact of server performance changeson energy consumption and the limitation of networkbandwidth In addition a local search operator is designedbased on which a two-layer genetic algorithm is proposede algorithm can schedule tens of thousands of tasks incloud and achieve large-scale optimization Yanggratoke

et al [85] proposed a general generic gossip protocol aimingat allocating resources in cloud environment An instanti-ation of this protocol was developed to enable server con-solidation to allocate resources to more servers to meetchanging load patterns

25 LoadBalancing Load balancing can help save energy bymanaging the number of servers and allocating resources

251 Cloud Computing Paya and Marinescu [86] intro-duced an operation model that balances cloud computingload and expands applications to save energy e principleof this model is to define an operating system e systemshould make servers run in the system as many as possible

Table 5 Work summary of scientific workflow execution in cloud computing

Problems Solutions Literatures Advantages

VM deployment A resource allocation method namedEnRealan [76] Performs scientific workflows based on energy perception

across cloud platforms

Workflowscheduling

A scheduling method based on energyperception [77]

Achieves a high parallelism without huge energyconsumption and minimizes the total consumption of

energy and execution time of workflows

A workflow scheduling method with severalobjects and hybrid particle swarm

optimization algorithm[78]

Makes the processors work at any voltage level minimizesthe energy consumption in the process of workflowscheduling and studies the scheduling problem of

workflows on heterogeneous systemsA scheduling algorithm based on various

applications [79] Enables service providers to gain high profits and reducesuser overhead at the same time

Cost reduction A scheduling algorithm based on energyperception [80 81] Minimizes the cost of performing workflows while

meeting the time constraintEffectiveimplementation

A flexible resource allocation and jobscheduling mechanism [82] Implements scientific workflows within prescribed

budgets and deadlines

Table 4 Work summary of resource allocation and management in IoT

Problems Solutions Literatures Advantages

Resourcemanagement

A distributed cloud network framework [66]Replaces the centralized architecture with a distributed cloud

architecture solves the defects of the centralized cloudarchitecture and brings people better experience

A MILP model [68] Assigns the requirement of servicesrsquo resource to heterogeneousnetwork equipment interface

A framework for communication used in5G [69]

Transforms the resource allocation problem into a power andchannel allocation problem minimizes the total energy

consumption and improves QoS levelsA low complexity channel allocation

algorithm [70] Improves throughput of the network and allocates finiteresources to a large group of users

A resource allocation framework withseveral bands under cognitive 5G IoT [71] Manages resources more flexibly and reduces energy

consumption than common single-band approach

A protocol which is distributed andoptimal to allocate resources [72]

Has excellent adaptability when changing topology of networkand dynamically manages resources in the heterogeneous IoT

environmentA multilevel allocating resourcesalgorithm for IoT communication [73] Has fast data processing rate and very low latency in both

saturated and nonsaturated environment

A new algorithm to allocate bandwidthdynamically [74]

Reduces the error of signal reconstruction under the samebandwidth and makes the bandwidth allocation of IoT more

reasonable

A reinforcement learning mechanism [75]Effectively solves the resource allocation problems caused by themismatch of service quality and complex service providing

condition in the IoT

Complexity 11

When no tasks are being performed the system should adjustservers to sleep thus energy consumption can be reducedJustafort et al [87] mainly studied the problem of workloaddistribution across cloud computing environment and pro-posed a method to solve the problem of the VM layout So thefootprint of carbon can be effectively reduced Panwar andMallick [88] proposed an algorithm to dynamically manage theload and effectively distribute the total incoming requestsbetween VMs rough efficient and uniform utilization ofresources this algorithm can achieve uniform distribution ofload between servers Yang et al [89] proposed a powermanagement mechanism to balance the load e system canmonitor VMs and dynamically allocate the resources Yanget al [90] proposed an optimization system to better allocateresources dynamically which can balance the load of VMsrunning on multiple physical machines Under this systemVMs can be migrated automatically to adjust high and lowloads without interrupting services Yang et al [89 90] manageVMs to achieve load balancing ey allocate resources dy-namically to migrate VMs which can balance workloads ondifferent physical machines e difference is that Yanget al[89] integrate a dynamic resource allocation approachwithOpenNebulaWhile Yang et al [90] focuse on avoiding serviceoutages during VM migration

Table 6 shows a summary of the abovementioned workse problems in Table 6 are from the load balancing problemServer management is about the control of the number ofservers running in the system Workload management is therational allocation of workload or tasks VMmanagement refersto configuringVMresources andmigratingVMs to adjust loads

252 Fog Computing Xu et al [91] proposed a methodcalled ldquoDRAMrdquo to dynamically allocate resources in fogcomputing environment which can avoid both too highand too low loads e method first analyzes the loadbalance of different kinds of computing nodes And then itdesigns a fog resource allocation method to achieve loadbalance which allocates resources statically and migratesservices dynamically Oueis et al [92] studied the loadbalance problem in fog computing A custom fog clus-tering algorithm is proposed to solve the problem In theproblem several users need to offload computations andall of their demands need to be handled by local com-puting cluster

253 IoT Wang et al [93] established architecture of theenergy saving targeted system in industrial IoT In additionin order to predict sleep intervals they developed a sleepscheduling and a wake protocol which provide a better wayfor energy saving

3 Conclusion

is paper did a comprehensive study of QoS optimi-zation and energy saving in cloud computing edgecomputing fog computing and IoT models We sum-marized five main problems and analyzed their solutionsproposed by existing works By conducting this surveywe aim to help readers have a deeper understanding onthe concepts of different computing models and thetechniques of QoS optimization and energy saving inthese models

e investigated papers focus on issues about ensuringQoS and reducing SLA violations and resource manage-ment In the case of QoS assurance and SLA violationsreduction the main solution of QoS assurance is efficientVM management is solution can meet customersrsquo re-quirements through reasonable scheduling and integrationof VMs Most of resource management techniques are re-alized by reasonable scheduling of resources which canreduce the waste of VMs servers and traffic

Disclosure

is manuscript is an extension of A Survey of QoS Opti-mization and Energy Saving in Cloud Edge and IoT in e9th EAI International Conference on Cloud Computing

Conflicts of Interest

e authors declare that there are no conflicts of interestregarding the publication of this paper

Acknowledgments

is work was supported by the National Natural ScienceFoundation of China (Grant no 61702274) Natural ScienceFoundation of Jiangsu Province (Grant no BK20170958)and PAPD

Table 6 Work summary of load balancing in cloud computing

Problems Solutions Literatures AdvantagesServermanagement

An operation model that can balance cloudcomputing load and expand application [86] Saves energy by managing the number of servers running in

the system

Workloadmanagement

A hybrid approach [87] Reduces the footprint of carbon and allocates workload acrosscloud computing

An algorithm to dynamically manage theload [88] Manages the load evens the load distribution between

servers and allocates tasks between VMs

VMmanagement

A green power administration mechanism [89] Monitors and jointly allocates VM resources

An optimization system [90]Migrates VMs to adjust high and low loads without

interrupting services and balances the load of VMs runningon multiple physical machines

12 Complexity

References

[1] Q Zhang L Cheng and R Boutaba ldquoCloud computingstate-of-the-art and research challengesrdquo Journal of InternetServices and Applications vol 1 no 1 pp 7ndash18 2010

[2] W Shi J Cao Q Zhang Y Li and L Xu ldquoEdge computingvision and challengesrdquo IEEE Internet of Aings Journal vol 3no 5 pp 637ndash646 2016

[3] M Iorga L Feldman R Barton M J Martin N S Gorenand C Mahmoudi ldquoFog computing conceptual modelrdquo TechRep Recommendations of the National Institute of Standardsand Technology Gaithersburg MD USA 2018

[4] Nebbiolo ldquoFog vs edge computingrdquo Tech Rep NebbioloTechnologies Inc Milpitas CA USA 2018

[5] C T Do N H Tran C Pham M G R Alam J H Son andC S Hong ldquoA proximal algorithm for joint resource allo-cation andminimizing carbon footprint in geo-distributed fogcomputingrdquo in Proceedings of the 2015 International Con-ference on Information Networking (ICOIN) pp 324ndash329IEEE Siem Reap Cambodia January 2015

[6] S Vashi J Ram J Modi S Verma and C Prakash ldquoInternetof things (IoT) a vision architectural elements and securityissuesrdquo in Proceedings of the 2017 International Conference onI-SMAC (IoT in Social Mobile Analytics and Cloud)(I-SMAC) pp 492ndash496 IEEE Coimbatore India February2017

[7] M Mazzucco D Dyachuk and R Deters ldquoMaximizing cloudprovidersrsquo revenues via energy aware allocation policiesrdquo inProceedings of the 2010 IEEE 3rd International Conference onCloud Computing pp 131ndash138 IEEE Miami FL USA July2010

[8] Q He R Zhou X Zhang et al ldquoKeyword search for buildingservice-based systemsrdquo IEEE Transactions on Software En-gineering vol 43 no 7 pp 658ndash674 2016

[9] L Sun H Dong O K Hussain F K Hussain and A X LiuldquoA framework of cloud service selection with criteria inter-actionsrdquo Future Generation Computer Systems vol 94pp 749ndash764 2019

[10] M Mazzucco and D Dyachuk ldquoOptimizing cloud providersrevenues via energy efficient server allocationrdquo SustainableComputing Informatics and Systems vol 2 no 1 pp 1ndash122012

[11] Q He J Han F Chen et al ldquoQos-aware service selection forcustomisable multi-tenant service-based systems maturityand approachesrdquo in Proceedings of the 2015 IEEE 8th Inter-national Conference on Cloud Computing pp 237ndash244 IEEENew York NY USA July 2015

[12] L Sun J Ma H Wang Y Zhang and J Yong ldquoCloud servicedescription model an extension of usdl for cloud servicesrdquoIEEE Transactions on Services Computing vol 11 no 2pp 354ndash368 2015

[13] Y Wang Q He D Ye and Y Yang ldquoFormulating criticality-based cost-effective fault tolerance strategies for multi-tenantservice-based systemsrdquo IEEE Transactions on Software En-gineering vol 44 no 3 pp 291ndash307 2017

[14] S Mustafa K Bilal S U R Malik and S A Madani ldquoSla-aware energy efficient resource management for cloud en-vironmentsrdquo IEEE Access vol 6 pp 15004ndash15020 2018

[15] J Bi H Yuan M Tie and W Tan ldquoSla-based optimisation ofvirtualised resource for multi-tier web applications in clouddata centresrdquo Enterprise Information Systems vol 9 no 7pp 743ndash767 2015

[16] S Singh I Chana and R Buyya ldquoStar sla-aware autonomicmanagement of cloud resourcesrdquo IEEE Transactions on CloudComputing p 1 2017

[17] A Beloglazov and R Buyya ldquoEnergy efficient resourcemanagement in virtualized cloud data centersrdquo in Proceedingsof the 2010 10th IEEEACM International Conference onCluster Cloud and Grid Computing pp 826ndash831 IEEEComputer Society Melbourne Australia May 2010

[18] M Guazzone C Anglano and M Canonico ldquoEnergy-effi-cient resource management for cloud computing infra-structuresrdquo in Proceedings of the 2011 IEEE AirdInternational Conference on Cloud Computing Technology andScience pp 424ndash431 IEEE Athens Greece November 2011

[19] Y Sun J White and S Eade ldquoA model-based system toautomate cloud resource allocation and optimizationrdquo inProceedings of the International Conference on Model DrivenEngineering Languages and Systems pp 18ndash34 SpringerValencia Spain October 2014

[20] G Siddesh and K Srinivasa ldquoSla-driven dynamic resourceallocation on cloudsrdquo in Proceedings of the InternationalConference on Advanced Computing Networking and Securitypp 9ndash18 Springer Surathkal India December 2011

[21] S K Garg S K Gopalaiyengar and R Buyya ldquoSla-basedresource provisioning for heterogeneous workloads in avirtualized cloud datacenterrdquo in Proceedings of the Interna-tional Conference on Algorithms and Architectures for ParallelProcessing pp 371ndash384 Springer Melbourne AustraliaOctober 2011

[22] J Bi Z Zhu and H Yuan ldquoSla-aware dynamic resourceprovisioning for profit maximization in shared cloud datacentersrdquo in Proceedings of the International Conference onHigh Performance Networking Computing and Communi-cation Systems pp 366ndash372 Springer Singapore May 2011

[23] L Qi Y Chen Y Yuan S Fu X Zhang and X Xu ldquoA Qos-aware virtual machine scheduling method for energy con-servation in cloud-based cyber-physical systemsrdquo WorldWide Web vol 4 no 3 pp 1ndash23 2019

[24] A Beloglazov J Abawajy and R Buyya ldquoEnergy-aware re-source allocation heuristics for efficient management of datacenters for cloud computingrdquo Future Generation ComputerSystems vol 28 no 5 pp 755ndash768 2012

[25] L Qi R Wang C Hu S Li Q He and X Xu ldquoTime-awaredistributed service recommendation with privacy-preserva-tionrdquo Information Sciences vol 480 pp 354ndash364 2019

[26] Z Zhai B Cheng Y Tian J Chen L Zhao and M Niu ldquoAdata-driven service creation approach for end-usersrdquo IEEEAccess vol 4 pp 9923ndash9940 2016

[27] L Gu D Zeng S Guo A Barnawi and Y Xiang ldquoCostefficient resource management in fog computing supportedmedical cyber-physical systemrdquo IEEE Transactions onEmerging Topics in Computing vol 5 no 1 pp 108ndash119 2015

[28] L Ni J Zhang C Jiang C Yan and K Yu ldquoResource al-location strategy in fog computing based on priced timed petrinetsrdquo Ieee Internet of Aings Journal vol 4 no 5pp 1216ndash1228 2017

[29] L Wei T Yang F C Delicato et al ldquoOn enabling sustainableedge computing with renewable energy resourcesrdquo IEEECommunications Magazine vol 56 no 5 pp 94ndash101 2018

[30] P Lai Q He G Cui et al ldquoEdge user allocation with dynamicquality of servicerdquo in Proceedings of the International Con-ference on Service-Oriented Computing pp 86ndash101 SpringerToulouse France October 2019

[31] X Xu Q Liu X Zhang J Zhang L Qi and W Dou ldquoAblockchain-powered crowdsourcing method with privacy

Complexity 13

preservation in mobile environmentrdquo IEEE Transactions onComputational Social Systems vol 6 no 6 pp 1407ndash1419 2019

[32] O Rolik E Zharikov and S Telenyk ldquoMicrocloud-basedarchitecture of management system for IoT infrastructuresrdquoin Proceedings of the 2016 Aird International Scientific-Practical Conference Problems of Infocommunications Scienceand Technology (PIC SampT) pp 149ndash151 IEEE KharkivUkaraine October 2016

[33] W He S Guo Y Liang R Ma X Qiu and L Shi ldquoQos-awareand resource-efficient dynamic slicingmechanism for internetof thingsrdquo Computers Materials amp Continua vol 61 no 3pp 1345ndash1364 2019

[34] J Yao and N Ansari ldquoQos-aware fog resource provisioningand mobile device power control in IoT networksrdquo IEEETransactions on Network and Service Management vol 16no 1 pp 167ndash175 2018

[35] W Wang Y Jiang and W Wu ldquoMultiagent-based resourceallocation for energy minimization in cloud computing sys-temsrdquo IEEE Transactions on Systems Man and CyberneticsSystems vol 47 no 2 pp 1ndash16 2016

[36] K Krishnajyothi ldquoParallel data processing for effective dy-namic resource allocation in the cloudrdquo International Journalof Computer Applications vol 70 no 22 pp 1ndash4 2013

[37] M M Hassan B Song M S Hossain and A Alamri ldquoEf-ficient resource scheduling for big data processing in cloudplatformrdquo in Proceedings of the International Conference onInternet and Distributed Computing Systems pp 51ndash63Springer Calabria Italy September 2014

[38] C-M Wu R-S Chang and H-Y Chan ldquoA green energy-efficient scheduling algorithm using the dvfs technique forcloud datacentersrdquo Future Generation Computer Systemsvol 37 no 7 pp 141ndash147 2014

[39] Z Wang and X Su ldquoDynamically hierarchical resource-al-location algorithm in cloud computing environmentrdquo AeJournal of Supercomputing vol 71 no 7 pp 2748ndash2766 2015

[40] W-Y Lin G-Y Lin and H-Y Wei ldquoDynamic auctionmechanism for cloud resource allocationrdquo in Proceedings ofthe 2010 10th IEEEACM International Conference on ClusterCloud and Grid Computing pp 591-592 IEEE ComputerSociety Victoria Australia May 2010

[41] Y O Yazir C Matthews R Farahbod et al ldquoDynamic re-source allocation in computing clouds using distributedmultiple criteria decision analysisrdquo in Proceedings of the 2010IEEE 3rd International Conference on Cloud Computingpp 91ndash98 IEEE Washington DC USA July 2010

[42] W Lin J Z Wang C Liang and D Qi ldquoA threshold-baseddynamic resource allocation scheme for cloud computingrdquoProcedia Engineering vol 23 no 5 pp 695ndash703 2011

[43] X Xu S Fu L Qi et al ldquoAn IoT-oriented data placementmethod with privacy preservation in cloud environmentrdquoJournal of Network and Computer Applications vol 124pp 148ndash157 2018

[44] J Bokyun P Md Jalil L Daeho and S Doug Young ldquoEf-ficient computation offloading in mobile cloud computing forvideo streaming over 5 grdquo Computers Materials amp Continuavol 61 no 2 pp 439ndash463 2019

[45] H Sarbazi-Azad and A Y Zomaya ldquoEnergy-efficient resourceutilization in cloud computingrdquo in Large Scale Network-Centric Distributed Systems pp 377ndash408 Wiley-IEEE PressHoboken NJ USA 2014

[46] C-H Hsu K D Slagter S-C Chen and Y-C ChungldquoOptimizing energy consumption with task consolidation incloudsrdquo Information Sciences vol 258 no 3 pp 452ndash4622014

[47] C-H Hsu S-C Chen C-C Lee et al ldquoEnergy-aware taskconsolidation technique for cloud computingrdquo in Proceedingsof the 2011 IEEE Aird International Conference on CloudComputing Technology and Science pp 115ndash121 IEEE Ath-ens Greece 2011

[48] S K Panda and P K Jana ldquoAn efficient task consolidationalgorithm for cloud computing systemsrdquo in Proceedings of theInternational Conference on Distributed Computing and In-ternet Technology pp 61ndash74 Springer Bhubaneswar IndiaJanuary 2016

[49] L Yin J Luo and H Luo ldquoTasks scheduling and resourceallocation in fog computing based on containers for smartmanufacturingrdquo IEEE Transactions on Industrial Informaticsvol 14 no 10 pp 4712ndash4721 2018

[50] M Aazam and E-N Huh ldquoDynamic resource provisioningthrough fog micro datacenterrdquo in Proceedings of the 2015IEEE International Conference on Pervasive Computing andCommunication Workshops (PerCom Workshops) pp 105ndash110 IEEE St Louis MO USA March 2015

[51] B Jia H Hu Y Zeng T Xu and Y Yang ldquoDouble-matchingresource allocation strategy in fog computing networks basedon cost efficiencyrdquo Journal of Communications and Networksvol 20 no 3 pp 237ndash246 2018

[52] H Zhang Y Xiao S Bu D Niyato F R Yu and Z HanldquoComputing resource allocation in three-tier IoT fog net-works a joint optimization approach combining stackelberggame and matchingrdquo IEEE Internet of Aings Journal vol 4no 5 pp 1204ndash1215 2017

[53] J Tan T-H Chang and T Q Quelc ldquoMinimum energyresource allocation in fog radio access network with fronthauland latency constraintsrdquo in Proceedings of the 2018 IEEE 19thInternational Workshop on Signal Processing Advances inWireless Communications (SPAWC) pp 1ndash5 IEEE KalamataGreece June 2018

[54] D R de Vasconcelos R M de Castro Andrade andJ N de Souza ldquoSmart shadowndashan autonomous availabilitycomputation resource allocation platform for internet ofthings in the fog computing environmentrdquo in Proceedings of the2015 International Conference on Distributed Computing inSensor Systems pp 216-217 IEEE Fortaleza Brazil June 2015

[55] M Aazam M St-Hilaire C-H Lung and I LambadarisldquoPre-fog IoT trace based probabilistic resource estimation atfogrdquo in Proceedings of the 2016 13th IEEE Annual ConsumerCommunications amp Networking Conference (CCNC) pp 12ndash17 IEEE Las Vegas NV USA January 2016

[56] N D Tung L L Bao and B Vijay ldquoPrice-based resourceallocation for edge computing a market equilibrium ap-proachrdquo IEEE Transactions on Cloud Computing p 1 2018

[57] X Xu C He Z Xu L Qi S Wan and M Z A BhuiyanldquoJoint optimization of offloading utility and privacy for edgecomputing enabled IoTrdquo IEEE Internet of Aings Journal2019

[58] X Xu Y Chen X Zhang Q Liu X Liu and L Qi ldquoAblockchain-based computation offloading method for edgecomputing in 5G networksrdquo Software Practice and Experi-ence 2019

[59] X Xu Y Li T Huang et al ldquoAn energy-aware computationoffloading method for smart edge computing in wirelessmetropolitan area networksrdquo Journal of Network and Com-puter Applications vol 133 pp 75ndash85 2019

[60] X Xu Y Xue L Qi et al ldquoAn edge computing-enabledcomputation offloading method with privacy preservation forinternet of connected vehiclesrdquo Future Generation ComputerSystems vol 96 pp 89ndash100 2019

14 Complexity

[61] G Yeting L Fang X Nong and C Zhengguo ldquoTask-basedresource allocation bid in edge computing micro datacenterrdquoComputers Materials amp Continua vol 61 no 2 pp 777ndash7922019

[62] X Chen L Jiao W Li and X Fu ldquoEfficient multi-usercomputation offloading for mobile-edge cloud computingrdquoIEEEACM Transactions on Networking vol 24 no 5pp 2795ndash2808 2016

[63] B Gao L He X Lu C Chang K Li and K Li ldquoDevelopingenergy-aware task allocation schemes in cloud-assisted mo-bile workflowsrdquo in Proceedings of the 2015 IEEE InternationalConference on Computer and Information Technology Ubiq-uitous Computing and Communications Dependable Auto-nomic and Secure Computing Pervasive Intelligence andComputing pp 1266ndash1273 IEEE Liverpool UK October2015

[64] X Xu X Zhang H Gao Y Xue L Qi andW Dou ldquoBecomeblockchain-enabled computation offloading for IoT in mobileedge computingrdquo IEEE Transactions on Industrial Infor-matics 2019

[65] W Yifei W Zhaoying G Da and Y F Richard ldquoDeepq-learning based computation offloading strategy for mobileedge computingrdquo Computers Materials amp Continua vol 59no 1 pp 89ndash104 2019

[66] M Barcelo A Correa J Llorca A M Tulino J L Vicarioand A Morell ldquoIoT-cloud service optimization in nextgeneration smart environmentsrdquo IEEE Journal on SelectedAreas in Communications vol 34 no 12 pp 4077ndash40902016

[67] B Cheng M Wang S Zhao Z Zhai D Zhu and J ChenldquoSituation-aware dynamic service coordination in an IoTenvironmentrdquo IEEEACM Transactions On Networkingvol 25 no 4 pp 2082ndash2095 2017

[68] V Angelakis I Avgouleas N Pappas and D Yuan ldquoFlexibleallocation of heterogeneous resources to services on an IoTdevicerdquo in Proceedings of the 2015 IEEE Conference onComputer Communications Workshops (INFOCOMWKSHPS) pp 99-100 IEEE Hong Kong China April 2015

[69] S Li Q Ni Y Sun G Min and S Al-Rubaye ldquoEnergy-ef-ficient resource allocation for industrial cyber-physical IoTsystems in 5g erardquo IEEE Transactions on Industrial Infor-matics vol 14 no 6 pp 2618ndash2628 2018

[70] X Liu Z Qin Y Gao and J A McCann ldquoResource allocationin wireless powered IoT networksrdquo IEEE Internet of AingsJournal vol 6 no 3 pp 4935ndash4945 2019

[71] W Ejaz and M Ibnkahla ldquoMulti-band spectrum sensing andresource allocation for IoT in cognitive 5g networksrdquo IEEEInternet of Aings Journal vol 5 no 1 pp 150ndash163 2017

[72] G Colistra V Pilloni and L Atzori ldquoTask allocation in groupof nodes in the IoT a consensus approachrdquo in Proceedings ofthe 2014 IEEE International Conference on Communications(ICC) pp 3848ndash3853 IEEE Sydney Australia June 2014

[73] J Li Q Sun and G Fan ldquoResource allocation for multiclassservice in IoTuplink communicationsrdquo inProceedings of the 20163rd International Conference on Systems and Informatics (ICSAI)pp 777ndash781 IEEE Shanghai China November 2016

[74] L I Zheng and K H Liu ldquoDynamic bandwidth resourceallocation algorithm in internet of things and its applicationrdquoComputer Engineering vol 38 no 17 pp 16ndash19 2012

[75] K Gai and M Qiu ldquoOptimal resource allocation using re-inforcement learning for IoT content-centric servicesrdquo Ap-plied Soft Computing vol 70 pp 12ndash21 2018

[76] X Xu W Dou X Zhang and J Chen ldquoEnreal an energy-aware resource allocation method for scientific workflow

executions in cloud environmentrdquo IEEE Transactions onCloud Computing vol 4 no 2 pp 166ndash179 2016

[77] K Bousselmi Z Brahmi and M M Gammoudi ldquoEnergyefficient partitioning and scheduling approach for scientificworkflows in the cloudrdquo in Proceedings of the 2016 IEEEInternational Conference on Services Computing (SCC)pp 146ndash154 IEEE San Francisco CA USA June 2016

[78] Y Sonia C Rachid K Hubert and G Bertrand ldquoMulti-objective approach for energy-aware workflow scheduling incloud computing environmentsrdquo Ae Scientific World Jour-nal vol 2013 Article ID 350934 13 pages 2013

[79] F Cao Efficient Scientific Workflow Scheduling in CloudEnvironment Southern Illinois University Carbondale ILUSA 2014

[80] Z Li J Ge H Hu W Song H Hu and B Luo ldquoCost andenergy aware scheduling algorithm for scientific workflowswith deadline constraint in cloudsrdquo IEEE Transactions onServices Computing vol 11 no 4 pp 713ndash726 2015

[81] M Khaleel and M M Zhu ldquoEnergy-aware job managementapproaches for workflow in cloudrdquo in Proceedings of the 2015IEEE International Conference on Cluster Computingpp 506-507 IEEE Chicago IL USA September 2015

[82] J Shi J Luo F Dong and J Zhang ldquoA budget and deadlineaware scientific workflow resource provisioning and sched-uling mechanism for cloudrdquo in Proceedings of the 2014 IEEE18th International Conference on Computer Supported Co-operative Work in Design (CSCWD) pp 672ndash677 IEEEHsinchu Taiwan January 2014

[83] Y Ge Y Zhang Q Qiu and Y-H Lu ldquoA game theoreticresource allocation for overall energy minimization in mobilecloud computing systemrdquo in Proceedings of the 2012ACMIEEE International Symposium on Low Power Electronicsand Design pp 279ndash284 ACM Redondo Beach CA USAAugust 2012

[84] X Wang Y Wang and C Yue ldquoAn energy-aware bi-leveloptimization model for multi-job scheduling problems undercloud computingrdquo Soft Computing vol 20 no 1 pp 303ndash3172016

[85] R Yanggratoke F Wuhib and R Stadler ldquoGossip-basedresource allocation for green computing in large cloudsrdquo inProceedings of the 2011 7th International Conference onNetwork and Service Management pp 1ndash9 IEEE ParisFrance October 2011

[86] A Paya and D C Marinescu ldquoEnergy-aware load balancingand application scaling for the cloud ecosystemrdquo IEEETransactions on Cloud Computing vol 5 no 1 pp 15ndash272015

[87] V D Justafort R Beaubrun and S Pierre ldquoA hybrid ap-proach for optimizing carbon footprint in intercloud envi-ronmentrdquo IEEE Transactions on Services Computing vol 12no 2 pp 186ndash198 2016

[88] R Panwar and B Mallick ldquoLoad balancing in cloud com-puting using dynamic load management algorithmrdquo inProceedings of the 2015 International Conference on GreenComputing and Internet of Aings (ICGCIoT) pp 773ndash778IEEE Noida India October 2015

[89] C-T Yang K-C Wang H-Y Cheng C-T Kuo andW C C Chu ldquoGreen power management with dynamicresource allocation for cloud virtual machinesrdquo in Proceedingsof the 2011 IEEE International Conference on High Perfor-mance Computing and Communications pp 726ndash733 IEEEAlberta Canada September 2011

[90] C-T Yang H-Y Cheng and K-L Huang ldquoA dynamicresource allocation model for virtual machine management

Complexity 15

on cloudrdquo in Proceedings of the International Conference onGrid and Distributed Computing pp 581ndash590 SpringerGangneug Korea December 2011

[91] X Xu F Shucun C Qing et al ldquoDynamic resource allocationfor load balancing in fog environmentrdquo Wireless Commu-nications amp Mobile Computing vol 2018 no 2 Article ID6421607 15 pages 2018

[92] J Oueis E C Strinati and S Barbarossa ldquoe fog balancingload distribution for small cell cloud computingrdquo in Pro-ceedings of the 2015 IEEE 81st Vehicular Technology Confer-ence (VTC Spring) pp 1ndash6 IEEE Glasgow UK May 2015

[93] K Wang Y Wang Y Sun S Guo and J Wu ldquoGreen in-dustrial internet of things architecture an energy-efficientperspectiverdquo IEEE Communications Magazine vol 54 no 12pp 48ndash54 2016

16 Complexity

resource allocation rather than for the whole service It reducesthe packet loss rate and saves energy by unloading services

224 MEC Chen et al [62] studied the problem of com-puting unloading with several users in the environment ofMEC with wireless interference which have many channels Inaddition a distributed algorithm for computing unloading isdevelopede algorithm can perform the unloading well evenwhen there are a large number of users Gao et al [63] built aquadratic binary program which is able to assign tasks inmobile cloud computing environment Two algorithms arepresented to obtain the optimal solution Both of these heuristicalgorithms can effectively solve the task assignment problemXu et al [64] proposed an offloadingmethod using block chaintechnology It can guarantee the loss of data in offloading tasksunder edge computing And it can solve the problem of re-source requests out of proportion due to the limited load of

edge computing equipment during task transfer Yifei et al [65]proposed a model-free reinforcement learning framework tosolve the problem of computational unloadingis model canbe applied to the computational unloading with time-changingcomputing requests

225 IoT Barcelo et al [66] expressed the problem ofservice allocation [67] as a mixed flow problem with min-imum cost which can be solved by LP solving this serviceallocation problem can solve the problems of unbalancednetwork load and delay of end-to-end service And it canalso figure out the problem of excessive consumption ofelectricity brought by the architecture of centralized cloudAngelakis et al [68] assigned the requirements of servicesresources to heterogeneous network interfaces of equipmentsSo more heterogeneous network interfaces can be used by alarge amount of services

Mobile devices

Mobile devices ECN

ECN

ECNECN

AP

AP

AP

APAP

Figure 5 An architecture of smart edge computing [59]

Table 3 Work summary of resource allocation and management in fog computing

Problems Solutions Literatures AdvantagesTaskallocation

A new model of scheduling jobs and aredistribution mechanism [49] Finishes jobs on time optimizes the number of tasks and

shortens the delay of tasks

Resourceallocation

A framework to administrate resources and amethod to predict and administrate resources [50]

Assists service providers to predict the amount of availableresources based on different types of service customers anddeals with the phenomenon that objects or devices withdraw

from resource utilization at any time

A bi-matching approach [51] Improves the performance of the system and obtains highercost efficiency

A framework for joint optimization [52] Optimizes resource allocation and improves networkperformance

A method to allocate computing andcommunication resources [53]

Allocates computing and communication resources transferscomputing jobs to remote cloud and nodes simplifies edge

nodesrsquo computing and saves computing energyA platform to allocate resources accessible to

client devices [54] Enables rapid response to computing resources and allocatesthe resources of devices near the host

A framework to manage resources [55] Considers the fluctuation of the customer abandonmentprobability and service types

Low latency A distributed algorithm [5] Solves the problem of wide range optimization and allocatesresources jointly

Complexity 9

Li et al [69] proposed communication framework in 5Gand studied the problem of allocating power andchannels So the signal data in the channel can beavailable and the total energy efficiency can be maximumFormula (9) [69] shows how to calculate the energy ef-ficiency of a system

U 1113944K

k11113944

M

i1EES

iK + 1113944K

K11113944

N

i1EEA

jk (9)

where EESiK and EEA

jk respectively denote the energy ef-ficiency of sensor S and actuator A on channels e sets ofsensors actuators and channels are respectively repre-sented as s S1 S2 SM1113864 1113865 A A1 A2 AN1113864 1113865 andC C1 C2 CK1113864 1113865

Liu et al [70] studied the problem of allocating resourcesefficiently on IoT that supplies wireless power In thismethod users are first grouped into accessible channelsAnd then power distribution of users grouped in the samechannel is studied to improve throughput of the networkis method can allocate finite resources to a large group ofusers Ejaz and Ibnkahla [71] proposed the resource allo-cation framework with several bands under cognitive 5GIoT In the highly dynamic environment of the IoTmultiband method can manage resources more flexiblyand reduce more energy consumption In addition areconstruction approach with several levels is proposed toallocate resources reasonably for applications with dif-ferent needs of QoS Colistra et al [72] proposed aprotocol which is distributed and optimal to allocateresources in heterogeneous IoT Because this protocol hasexcellent adaptability when changing topology of net-work it can distribute resources evenly among nodes Jianet al [73] proposed a multilevel allocating resources al-gorithm for IoT communication using advanced tech-nology e algorithm uses hierarchical structure and hasfast data processing rate and very low latency in a satu-rated or not saturated environment

Zheng and Liu [74] proposed a new algorithm to allocatebandwidth dynamically for controlling remote computers inthe IoT is method can reduce the error of signal re-construction under the same bandwidth and make thebandwidth allocation of IoT more reasonable Gai and Qiu[75] used reinforcement learning mechanisms to allocateresources to achieve high Quality of Experience ismethod can effectively solve the resource allocation prob-lems caused by the mismatch of service quality and complexservice providing condition in the IoT

Table 4 shows a summary of the above works eproblem in Table 4 represents the realization of dynamic andflexible allocation of resources e resources here canrepresent channels bandwidth and power

23 Scientific Workflow Execution Implementing scientificworkflows especially in heterogeneous environments canreduce resource waste and reduce energy costs Scientificworkflow can be obtained by reasonably allocating resourcesand dynamically deploying VMs

231 Cloud Computing Xu et al [76] proposed a resourceallocation method called EnRealan to solve the problem ofenergy consumption e dynamic deployment of VMs isgenerally adopted to execute scientific workflows Bousselmiet al [77] proposed a scheduling method based on energyperception for executing scientific workflows in cloudcomputing At first an algorithm of splitting workflows forenergy minimization is presented which can achieve a highparallelism without huge energy consumption en aheuristic algorithm used to optimize cat swarm is proposedfor the created partitions e algorithm can minimize thetotal consumption of energy and the execution time ofworkflows Sonia et al [78] proposed a workflow schedulingmethod with several objects and hybrid particle swarmoptimization algorithm In addition a method for dy-namically scaling voltage and frequency is proposed emethod canmake the processors work at any voltage level soas to minimize the energy consumption in the process ofworkflow scheduling Both Bousselmi et al [77] and Soniaet al [78] use scheduling method to achieve scientificworkflows and study the problem of energy consumptione difference is that Bousselmi et al [77] focuses on in-tensive computing tasks while Sonia et al [78] focuses onworkflow scheduling on heterogeneous computing systems

Cao [79] established a scheduling algorithm of scientificworkflows with an objective of energy savingis algorithmcan enable service providers to gain high profits and reduceusersrsquo overhead at the same time Li et al [80] proposed ascheduling algorithm based on cloud computing which canminimize cost of performing workflows within a specifiedtime In addition the rented VM was modified to save costfurther Khaleel and Zhu [81] proposed a scheduling algo-rithm and took scientific workflows as a model to make fulluse of cloud resources and save energy Shi et al [82]designed a flexible resource allocation and job schedulingmechanism to implement scientific workflows Because thismechanism can implement scientific workflows withinprescribed budgets and deadlines so it can work better thanother mechanisms

Table 5 shows a summary of the abovementioned workse problems in Table 5 are derived from the imple-mentation of the scientific workflow VM deployment refersto the rational allocation of VMs Workflow scheduling refersto reducing the scheduling energy and time In addition itrefers to the scheduling of the workflow on heterogeneoussystems Cost reduction refers to reducing the cost of workflowexecution Effective implementation is about scientific work-flow execution within a specified budget and time

24 Server Optimization Server optimization is also a goodway to save energy e goal of optimizing the server can beachieved by uninstalling unnecessary servers or consoli-dating servers as well as by reasonably scheduling tasksUnlike QoS optimization server optimization aims to op-timize the number of used servers improves the energyefficiency of servers and consolidates servers However QoSoptimization studies how tomake users get better experienceand meet their needs

10 Complexity

241 Cloud Computing Ge et al [83] proposed a game-theoretic method and transformed the problem of mini-mizing energy into a congestion game All mobile devices inthis method are participants in the game e methodchooses a server to unload the computation tasks to optimizeQoS levels and save energy which can optimize the systemand save energy Wang et al [84] proposed a MapReduce-based multitask scheduling algorithm to achieve the ob-jective of energy saving is model is a two-layer modelwhich considers the impact of server performance changeson energy consumption and the limitation of networkbandwidth In addition a local search operator is designedbased on which a two-layer genetic algorithm is proposede algorithm can schedule tens of thousands of tasks incloud and achieve large-scale optimization Yanggratoke

et al [85] proposed a general generic gossip protocol aimingat allocating resources in cloud environment An instanti-ation of this protocol was developed to enable server con-solidation to allocate resources to more servers to meetchanging load patterns

25 LoadBalancing Load balancing can help save energy bymanaging the number of servers and allocating resources

251 Cloud Computing Paya and Marinescu [86] intro-duced an operation model that balances cloud computingload and expands applications to save energy e principleof this model is to define an operating system e systemshould make servers run in the system as many as possible

Table 5 Work summary of scientific workflow execution in cloud computing

Problems Solutions Literatures Advantages

VM deployment A resource allocation method namedEnRealan [76] Performs scientific workflows based on energy perception

across cloud platforms

Workflowscheduling

A scheduling method based on energyperception [77]

Achieves a high parallelism without huge energyconsumption and minimizes the total consumption of

energy and execution time of workflows

A workflow scheduling method with severalobjects and hybrid particle swarm

optimization algorithm[78]

Makes the processors work at any voltage level minimizesthe energy consumption in the process of workflowscheduling and studies the scheduling problem of

workflows on heterogeneous systemsA scheduling algorithm based on various

applications [79] Enables service providers to gain high profits and reducesuser overhead at the same time

Cost reduction A scheduling algorithm based on energyperception [80 81] Minimizes the cost of performing workflows while

meeting the time constraintEffectiveimplementation

A flexible resource allocation and jobscheduling mechanism [82] Implements scientific workflows within prescribed

budgets and deadlines

Table 4 Work summary of resource allocation and management in IoT

Problems Solutions Literatures Advantages

Resourcemanagement

A distributed cloud network framework [66]Replaces the centralized architecture with a distributed cloud

architecture solves the defects of the centralized cloudarchitecture and brings people better experience

A MILP model [68] Assigns the requirement of servicesrsquo resource to heterogeneousnetwork equipment interface

A framework for communication used in5G [69]

Transforms the resource allocation problem into a power andchannel allocation problem minimizes the total energy

consumption and improves QoS levelsA low complexity channel allocation

algorithm [70] Improves throughput of the network and allocates finiteresources to a large group of users

A resource allocation framework withseveral bands under cognitive 5G IoT [71] Manages resources more flexibly and reduces energy

consumption than common single-band approach

A protocol which is distributed andoptimal to allocate resources [72]

Has excellent adaptability when changing topology of networkand dynamically manages resources in the heterogeneous IoT

environmentA multilevel allocating resourcesalgorithm for IoT communication [73] Has fast data processing rate and very low latency in both

saturated and nonsaturated environment

A new algorithm to allocate bandwidthdynamically [74]

Reduces the error of signal reconstruction under the samebandwidth and makes the bandwidth allocation of IoT more

reasonable

A reinforcement learning mechanism [75]Effectively solves the resource allocation problems caused by themismatch of service quality and complex service providing

condition in the IoT

Complexity 11

When no tasks are being performed the system should adjustservers to sleep thus energy consumption can be reducedJustafort et al [87] mainly studied the problem of workloaddistribution across cloud computing environment and pro-posed a method to solve the problem of the VM layout So thefootprint of carbon can be effectively reduced Panwar andMallick [88] proposed an algorithm to dynamically manage theload and effectively distribute the total incoming requestsbetween VMs rough efficient and uniform utilization ofresources this algorithm can achieve uniform distribution ofload between servers Yang et al [89] proposed a powermanagement mechanism to balance the load e system canmonitor VMs and dynamically allocate the resources Yanget al [90] proposed an optimization system to better allocateresources dynamically which can balance the load of VMsrunning on multiple physical machines Under this systemVMs can be migrated automatically to adjust high and lowloads without interrupting services Yang et al [89 90] manageVMs to achieve load balancing ey allocate resources dy-namically to migrate VMs which can balance workloads ondifferent physical machines e difference is that Yanget al[89] integrate a dynamic resource allocation approachwithOpenNebulaWhile Yang et al [90] focuse on avoiding serviceoutages during VM migration

Table 6 shows a summary of the abovementioned workse problems in Table 6 are from the load balancing problemServer management is about the control of the number ofservers running in the system Workload management is therational allocation of workload or tasks VMmanagement refersto configuringVMresources andmigratingVMs to adjust loads

252 Fog Computing Xu et al [91] proposed a methodcalled ldquoDRAMrdquo to dynamically allocate resources in fogcomputing environment which can avoid both too highand too low loads e method first analyzes the loadbalance of different kinds of computing nodes And then itdesigns a fog resource allocation method to achieve loadbalance which allocates resources statically and migratesservices dynamically Oueis et al [92] studied the loadbalance problem in fog computing A custom fog clus-tering algorithm is proposed to solve the problem In theproblem several users need to offload computations andall of their demands need to be handled by local com-puting cluster

253 IoT Wang et al [93] established architecture of theenergy saving targeted system in industrial IoT In additionin order to predict sleep intervals they developed a sleepscheduling and a wake protocol which provide a better wayfor energy saving

3 Conclusion

is paper did a comprehensive study of QoS optimi-zation and energy saving in cloud computing edgecomputing fog computing and IoT models We sum-marized five main problems and analyzed their solutionsproposed by existing works By conducting this surveywe aim to help readers have a deeper understanding onthe concepts of different computing models and thetechniques of QoS optimization and energy saving inthese models

e investigated papers focus on issues about ensuringQoS and reducing SLA violations and resource manage-ment In the case of QoS assurance and SLA violationsreduction the main solution of QoS assurance is efficientVM management is solution can meet customersrsquo re-quirements through reasonable scheduling and integrationof VMs Most of resource management techniques are re-alized by reasonable scheduling of resources which canreduce the waste of VMs servers and traffic

Disclosure

is manuscript is an extension of A Survey of QoS Opti-mization and Energy Saving in Cloud Edge and IoT in e9th EAI International Conference on Cloud Computing

Conflicts of Interest

e authors declare that there are no conflicts of interestregarding the publication of this paper

Acknowledgments

is work was supported by the National Natural ScienceFoundation of China (Grant no 61702274) Natural ScienceFoundation of Jiangsu Province (Grant no BK20170958)and PAPD

Table 6 Work summary of load balancing in cloud computing

Problems Solutions Literatures AdvantagesServermanagement

An operation model that can balance cloudcomputing load and expand application [86] Saves energy by managing the number of servers running in

the system

Workloadmanagement

A hybrid approach [87] Reduces the footprint of carbon and allocates workload acrosscloud computing

An algorithm to dynamically manage theload [88] Manages the load evens the load distribution between

servers and allocates tasks between VMs

VMmanagement

A green power administration mechanism [89] Monitors and jointly allocates VM resources

An optimization system [90]Migrates VMs to adjust high and low loads without

interrupting services and balances the load of VMs runningon multiple physical machines

12 Complexity

References

[1] Q Zhang L Cheng and R Boutaba ldquoCloud computingstate-of-the-art and research challengesrdquo Journal of InternetServices and Applications vol 1 no 1 pp 7ndash18 2010

[2] W Shi J Cao Q Zhang Y Li and L Xu ldquoEdge computingvision and challengesrdquo IEEE Internet of Aings Journal vol 3no 5 pp 637ndash646 2016

[3] M Iorga L Feldman R Barton M J Martin N S Gorenand C Mahmoudi ldquoFog computing conceptual modelrdquo TechRep Recommendations of the National Institute of Standardsand Technology Gaithersburg MD USA 2018

[4] Nebbiolo ldquoFog vs edge computingrdquo Tech Rep NebbioloTechnologies Inc Milpitas CA USA 2018

[5] C T Do N H Tran C Pham M G R Alam J H Son andC S Hong ldquoA proximal algorithm for joint resource allo-cation andminimizing carbon footprint in geo-distributed fogcomputingrdquo in Proceedings of the 2015 International Con-ference on Information Networking (ICOIN) pp 324ndash329IEEE Siem Reap Cambodia January 2015

[6] S Vashi J Ram J Modi S Verma and C Prakash ldquoInternetof things (IoT) a vision architectural elements and securityissuesrdquo in Proceedings of the 2017 International Conference onI-SMAC (IoT in Social Mobile Analytics and Cloud)(I-SMAC) pp 492ndash496 IEEE Coimbatore India February2017

[7] M Mazzucco D Dyachuk and R Deters ldquoMaximizing cloudprovidersrsquo revenues via energy aware allocation policiesrdquo inProceedings of the 2010 IEEE 3rd International Conference onCloud Computing pp 131ndash138 IEEE Miami FL USA July2010

[8] Q He R Zhou X Zhang et al ldquoKeyword search for buildingservice-based systemsrdquo IEEE Transactions on Software En-gineering vol 43 no 7 pp 658ndash674 2016

[9] L Sun H Dong O K Hussain F K Hussain and A X LiuldquoA framework of cloud service selection with criteria inter-actionsrdquo Future Generation Computer Systems vol 94pp 749ndash764 2019

[10] M Mazzucco and D Dyachuk ldquoOptimizing cloud providersrevenues via energy efficient server allocationrdquo SustainableComputing Informatics and Systems vol 2 no 1 pp 1ndash122012

[11] Q He J Han F Chen et al ldquoQos-aware service selection forcustomisable multi-tenant service-based systems maturityand approachesrdquo in Proceedings of the 2015 IEEE 8th Inter-national Conference on Cloud Computing pp 237ndash244 IEEENew York NY USA July 2015

[12] L Sun J Ma H Wang Y Zhang and J Yong ldquoCloud servicedescription model an extension of usdl for cloud servicesrdquoIEEE Transactions on Services Computing vol 11 no 2pp 354ndash368 2015

[13] Y Wang Q He D Ye and Y Yang ldquoFormulating criticality-based cost-effective fault tolerance strategies for multi-tenantservice-based systemsrdquo IEEE Transactions on Software En-gineering vol 44 no 3 pp 291ndash307 2017

[14] S Mustafa K Bilal S U R Malik and S A Madani ldquoSla-aware energy efficient resource management for cloud en-vironmentsrdquo IEEE Access vol 6 pp 15004ndash15020 2018

[15] J Bi H Yuan M Tie and W Tan ldquoSla-based optimisation ofvirtualised resource for multi-tier web applications in clouddata centresrdquo Enterprise Information Systems vol 9 no 7pp 743ndash767 2015

[16] S Singh I Chana and R Buyya ldquoStar sla-aware autonomicmanagement of cloud resourcesrdquo IEEE Transactions on CloudComputing p 1 2017

[17] A Beloglazov and R Buyya ldquoEnergy efficient resourcemanagement in virtualized cloud data centersrdquo in Proceedingsof the 2010 10th IEEEACM International Conference onCluster Cloud and Grid Computing pp 826ndash831 IEEEComputer Society Melbourne Australia May 2010

[18] M Guazzone C Anglano and M Canonico ldquoEnergy-effi-cient resource management for cloud computing infra-structuresrdquo in Proceedings of the 2011 IEEE AirdInternational Conference on Cloud Computing Technology andScience pp 424ndash431 IEEE Athens Greece November 2011

[19] Y Sun J White and S Eade ldquoA model-based system toautomate cloud resource allocation and optimizationrdquo inProceedings of the International Conference on Model DrivenEngineering Languages and Systems pp 18ndash34 SpringerValencia Spain October 2014

[20] G Siddesh and K Srinivasa ldquoSla-driven dynamic resourceallocation on cloudsrdquo in Proceedings of the InternationalConference on Advanced Computing Networking and Securitypp 9ndash18 Springer Surathkal India December 2011

[21] S K Garg S K Gopalaiyengar and R Buyya ldquoSla-basedresource provisioning for heterogeneous workloads in avirtualized cloud datacenterrdquo in Proceedings of the Interna-tional Conference on Algorithms and Architectures for ParallelProcessing pp 371ndash384 Springer Melbourne AustraliaOctober 2011

[22] J Bi Z Zhu and H Yuan ldquoSla-aware dynamic resourceprovisioning for profit maximization in shared cloud datacentersrdquo in Proceedings of the International Conference onHigh Performance Networking Computing and Communi-cation Systems pp 366ndash372 Springer Singapore May 2011

[23] L Qi Y Chen Y Yuan S Fu X Zhang and X Xu ldquoA Qos-aware virtual machine scheduling method for energy con-servation in cloud-based cyber-physical systemsrdquo WorldWide Web vol 4 no 3 pp 1ndash23 2019

[24] A Beloglazov J Abawajy and R Buyya ldquoEnergy-aware re-source allocation heuristics for efficient management of datacenters for cloud computingrdquo Future Generation ComputerSystems vol 28 no 5 pp 755ndash768 2012

[25] L Qi R Wang C Hu S Li Q He and X Xu ldquoTime-awaredistributed service recommendation with privacy-preserva-tionrdquo Information Sciences vol 480 pp 354ndash364 2019

[26] Z Zhai B Cheng Y Tian J Chen L Zhao and M Niu ldquoAdata-driven service creation approach for end-usersrdquo IEEEAccess vol 4 pp 9923ndash9940 2016

[27] L Gu D Zeng S Guo A Barnawi and Y Xiang ldquoCostefficient resource management in fog computing supportedmedical cyber-physical systemrdquo IEEE Transactions onEmerging Topics in Computing vol 5 no 1 pp 108ndash119 2015

[28] L Ni J Zhang C Jiang C Yan and K Yu ldquoResource al-location strategy in fog computing based on priced timed petrinetsrdquo Ieee Internet of Aings Journal vol 4 no 5pp 1216ndash1228 2017

[29] L Wei T Yang F C Delicato et al ldquoOn enabling sustainableedge computing with renewable energy resourcesrdquo IEEECommunications Magazine vol 56 no 5 pp 94ndash101 2018

[30] P Lai Q He G Cui et al ldquoEdge user allocation with dynamicquality of servicerdquo in Proceedings of the International Con-ference on Service-Oriented Computing pp 86ndash101 SpringerToulouse France October 2019

[31] X Xu Q Liu X Zhang J Zhang L Qi and W Dou ldquoAblockchain-powered crowdsourcing method with privacy

Complexity 13

preservation in mobile environmentrdquo IEEE Transactions onComputational Social Systems vol 6 no 6 pp 1407ndash1419 2019

[32] O Rolik E Zharikov and S Telenyk ldquoMicrocloud-basedarchitecture of management system for IoT infrastructuresrdquoin Proceedings of the 2016 Aird International Scientific-Practical Conference Problems of Infocommunications Scienceand Technology (PIC SampT) pp 149ndash151 IEEE KharkivUkaraine October 2016

[33] W He S Guo Y Liang R Ma X Qiu and L Shi ldquoQos-awareand resource-efficient dynamic slicingmechanism for internetof thingsrdquo Computers Materials amp Continua vol 61 no 3pp 1345ndash1364 2019

[34] J Yao and N Ansari ldquoQos-aware fog resource provisioningand mobile device power control in IoT networksrdquo IEEETransactions on Network and Service Management vol 16no 1 pp 167ndash175 2018

[35] W Wang Y Jiang and W Wu ldquoMultiagent-based resourceallocation for energy minimization in cloud computing sys-temsrdquo IEEE Transactions on Systems Man and CyberneticsSystems vol 47 no 2 pp 1ndash16 2016

[36] K Krishnajyothi ldquoParallel data processing for effective dy-namic resource allocation in the cloudrdquo International Journalof Computer Applications vol 70 no 22 pp 1ndash4 2013

[37] M M Hassan B Song M S Hossain and A Alamri ldquoEf-ficient resource scheduling for big data processing in cloudplatformrdquo in Proceedings of the International Conference onInternet and Distributed Computing Systems pp 51ndash63Springer Calabria Italy September 2014

[38] C-M Wu R-S Chang and H-Y Chan ldquoA green energy-efficient scheduling algorithm using the dvfs technique forcloud datacentersrdquo Future Generation Computer Systemsvol 37 no 7 pp 141ndash147 2014

[39] Z Wang and X Su ldquoDynamically hierarchical resource-al-location algorithm in cloud computing environmentrdquo AeJournal of Supercomputing vol 71 no 7 pp 2748ndash2766 2015

[40] W-Y Lin G-Y Lin and H-Y Wei ldquoDynamic auctionmechanism for cloud resource allocationrdquo in Proceedings ofthe 2010 10th IEEEACM International Conference on ClusterCloud and Grid Computing pp 591-592 IEEE ComputerSociety Victoria Australia May 2010

[41] Y O Yazir C Matthews R Farahbod et al ldquoDynamic re-source allocation in computing clouds using distributedmultiple criteria decision analysisrdquo in Proceedings of the 2010IEEE 3rd International Conference on Cloud Computingpp 91ndash98 IEEE Washington DC USA July 2010

[42] W Lin J Z Wang C Liang and D Qi ldquoA threshold-baseddynamic resource allocation scheme for cloud computingrdquoProcedia Engineering vol 23 no 5 pp 695ndash703 2011

[43] X Xu S Fu L Qi et al ldquoAn IoT-oriented data placementmethod with privacy preservation in cloud environmentrdquoJournal of Network and Computer Applications vol 124pp 148ndash157 2018

[44] J Bokyun P Md Jalil L Daeho and S Doug Young ldquoEf-ficient computation offloading in mobile cloud computing forvideo streaming over 5 grdquo Computers Materials amp Continuavol 61 no 2 pp 439ndash463 2019

[45] H Sarbazi-Azad and A Y Zomaya ldquoEnergy-efficient resourceutilization in cloud computingrdquo in Large Scale Network-Centric Distributed Systems pp 377ndash408 Wiley-IEEE PressHoboken NJ USA 2014

[46] C-H Hsu K D Slagter S-C Chen and Y-C ChungldquoOptimizing energy consumption with task consolidation incloudsrdquo Information Sciences vol 258 no 3 pp 452ndash4622014

[47] C-H Hsu S-C Chen C-C Lee et al ldquoEnergy-aware taskconsolidation technique for cloud computingrdquo in Proceedingsof the 2011 IEEE Aird International Conference on CloudComputing Technology and Science pp 115ndash121 IEEE Ath-ens Greece 2011

[48] S K Panda and P K Jana ldquoAn efficient task consolidationalgorithm for cloud computing systemsrdquo in Proceedings of theInternational Conference on Distributed Computing and In-ternet Technology pp 61ndash74 Springer Bhubaneswar IndiaJanuary 2016

[49] L Yin J Luo and H Luo ldquoTasks scheduling and resourceallocation in fog computing based on containers for smartmanufacturingrdquo IEEE Transactions on Industrial Informaticsvol 14 no 10 pp 4712ndash4721 2018

[50] M Aazam and E-N Huh ldquoDynamic resource provisioningthrough fog micro datacenterrdquo in Proceedings of the 2015IEEE International Conference on Pervasive Computing andCommunication Workshops (PerCom Workshops) pp 105ndash110 IEEE St Louis MO USA March 2015

[51] B Jia H Hu Y Zeng T Xu and Y Yang ldquoDouble-matchingresource allocation strategy in fog computing networks basedon cost efficiencyrdquo Journal of Communications and Networksvol 20 no 3 pp 237ndash246 2018

[52] H Zhang Y Xiao S Bu D Niyato F R Yu and Z HanldquoComputing resource allocation in three-tier IoT fog net-works a joint optimization approach combining stackelberggame and matchingrdquo IEEE Internet of Aings Journal vol 4no 5 pp 1204ndash1215 2017

[53] J Tan T-H Chang and T Q Quelc ldquoMinimum energyresource allocation in fog radio access network with fronthauland latency constraintsrdquo in Proceedings of the 2018 IEEE 19thInternational Workshop on Signal Processing Advances inWireless Communications (SPAWC) pp 1ndash5 IEEE KalamataGreece June 2018

[54] D R de Vasconcelos R M de Castro Andrade andJ N de Souza ldquoSmart shadowndashan autonomous availabilitycomputation resource allocation platform for internet ofthings in the fog computing environmentrdquo in Proceedings of the2015 International Conference on Distributed Computing inSensor Systems pp 216-217 IEEE Fortaleza Brazil June 2015

[55] M Aazam M St-Hilaire C-H Lung and I LambadarisldquoPre-fog IoT trace based probabilistic resource estimation atfogrdquo in Proceedings of the 2016 13th IEEE Annual ConsumerCommunications amp Networking Conference (CCNC) pp 12ndash17 IEEE Las Vegas NV USA January 2016

[56] N D Tung L L Bao and B Vijay ldquoPrice-based resourceallocation for edge computing a market equilibrium ap-proachrdquo IEEE Transactions on Cloud Computing p 1 2018

[57] X Xu C He Z Xu L Qi S Wan and M Z A BhuiyanldquoJoint optimization of offloading utility and privacy for edgecomputing enabled IoTrdquo IEEE Internet of Aings Journal2019

[58] X Xu Y Chen X Zhang Q Liu X Liu and L Qi ldquoAblockchain-based computation offloading method for edgecomputing in 5G networksrdquo Software Practice and Experi-ence 2019

[59] X Xu Y Li T Huang et al ldquoAn energy-aware computationoffloading method for smart edge computing in wirelessmetropolitan area networksrdquo Journal of Network and Com-puter Applications vol 133 pp 75ndash85 2019

[60] X Xu Y Xue L Qi et al ldquoAn edge computing-enabledcomputation offloading method with privacy preservation forinternet of connected vehiclesrdquo Future Generation ComputerSystems vol 96 pp 89ndash100 2019

14 Complexity

[61] G Yeting L Fang X Nong and C Zhengguo ldquoTask-basedresource allocation bid in edge computing micro datacenterrdquoComputers Materials amp Continua vol 61 no 2 pp 777ndash7922019

[62] X Chen L Jiao W Li and X Fu ldquoEfficient multi-usercomputation offloading for mobile-edge cloud computingrdquoIEEEACM Transactions on Networking vol 24 no 5pp 2795ndash2808 2016

[63] B Gao L He X Lu C Chang K Li and K Li ldquoDevelopingenergy-aware task allocation schemes in cloud-assisted mo-bile workflowsrdquo in Proceedings of the 2015 IEEE InternationalConference on Computer and Information Technology Ubiq-uitous Computing and Communications Dependable Auto-nomic and Secure Computing Pervasive Intelligence andComputing pp 1266ndash1273 IEEE Liverpool UK October2015

[64] X Xu X Zhang H Gao Y Xue L Qi andW Dou ldquoBecomeblockchain-enabled computation offloading for IoT in mobileedge computingrdquo IEEE Transactions on Industrial Infor-matics 2019

[65] W Yifei W Zhaoying G Da and Y F Richard ldquoDeepq-learning based computation offloading strategy for mobileedge computingrdquo Computers Materials amp Continua vol 59no 1 pp 89ndash104 2019

[66] M Barcelo A Correa J Llorca A M Tulino J L Vicarioand A Morell ldquoIoT-cloud service optimization in nextgeneration smart environmentsrdquo IEEE Journal on SelectedAreas in Communications vol 34 no 12 pp 4077ndash40902016

[67] B Cheng M Wang S Zhao Z Zhai D Zhu and J ChenldquoSituation-aware dynamic service coordination in an IoTenvironmentrdquo IEEEACM Transactions On Networkingvol 25 no 4 pp 2082ndash2095 2017

[68] V Angelakis I Avgouleas N Pappas and D Yuan ldquoFlexibleallocation of heterogeneous resources to services on an IoTdevicerdquo in Proceedings of the 2015 IEEE Conference onComputer Communications Workshops (INFOCOMWKSHPS) pp 99-100 IEEE Hong Kong China April 2015

[69] S Li Q Ni Y Sun G Min and S Al-Rubaye ldquoEnergy-ef-ficient resource allocation for industrial cyber-physical IoTsystems in 5g erardquo IEEE Transactions on Industrial Infor-matics vol 14 no 6 pp 2618ndash2628 2018

[70] X Liu Z Qin Y Gao and J A McCann ldquoResource allocationin wireless powered IoT networksrdquo IEEE Internet of AingsJournal vol 6 no 3 pp 4935ndash4945 2019

[71] W Ejaz and M Ibnkahla ldquoMulti-band spectrum sensing andresource allocation for IoT in cognitive 5g networksrdquo IEEEInternet of Aings Journal vol 5 no 1 pp 150ndash163 2017

[72] G Colistra V Pilloni and L Atzori ldquoTask allocation in groupof nodes in the IoT a consensus approachrdquo in Proceedings ofthe 2014 IEEE International Conference on Communications(ICC) pp 3848ndash3853 IEEE Sydney Australia June 2014

[73] J Li Q Sun and G Fan ldquoResource allocation for multiclassservice in IoTuplink communicationsrdquo inProceedings of the 20163rd International Conference on Systems and Informatics (ICSAI)pp 777ndash781 IEEE Shanghai China November 2016

[74] L I Zheng and K H Liu ldquoDynamic bandwidth resourceallocation algorithm in internet of things and its applicationrdquoComputer Engineering vol 38 no 17 pp 16ndash19 2012

[75] K Gai and M Qiu ldquoOptimal resource allocation using re-inforcement learning for IoT content-centric servicesrdquo Ap-plied Soft Computing vol 70 pp 12ndash21 2018

[76] X Xu W Dou X Zhang and J Chen ldquoEnreal an energy-aware resource allocation method for scientific workflow

executions in cloud environmentrdquo IEEE Transactions onCloud Computing vol 4 no 2 pp 166ndash179 2016

[77] K Bousselmi Z Brahmi and M M Gammoudi ldquoEnergyefficient partitioning and scheduling approach for scientificworkflows in the cloudrdquo in Proceedings of the 2016 IEEEInternational Conference on Services Computing (SCC)pp 146ndash154 IEEE San Francisco CA USA June 2016

[78] Y Sonia C Rachid K Hubert and G Bertrand ldquoMulti-objective approach for energy-aware workflow scheduling incloud computing environmentsrdquo Ae Scientific World Jour-nal vol 2013 Article ID 350934 13 pages 2013

[79] F Cao Efficient Scientific Workflow Scheduling in CloudEnvironment Southern Illinois University Carbondale ILUSA 2014

[80] Z Li J Ge H Hu W Song H Hu and B Luo ldquoCost andenergy aware scheduling algorithm for scientific workflowswith deadline constraint in cloudsrdquo IEEE Transactions onServices Computing vol 11 no 4 pp 713ndash726 2015

[81] M Khaleel and M M Zhu ldquoEnergy-aware job managementapproaches for workflow in cloudrdquo in Proceedings of the 2015IEEE International Conference on Cluster Computingpp 506-507 IEEE Chicago IL USA September 2015

[82] J Shi J Luo F Dong and J Zhang ldquoA budget and deadlineaware scientific workflow resource provisioning and sched-uling mechanism for cloudrdquo in Proceedings of the 2014 IEEE18th International Conference on Computer Supported Co-operative Work in Design (CSCWD) pp 672ndash677 IEEEHsinchu Taiwan January 2014

[83] Y Ge Y Zhang Q Qiu and Y-H Lu ldquoA game theoreticresource allocation for overall energy minimization in mobilecloud computing systemrdquo in Proceedings of the 2012ACMIEEE International Symposium on Low Power Electronicsand Design pp 279ndash284 ACM Redondo Beach CA USAAugust 2012

[84] X Wang Y Wang and C Yue ldquoAn energy-aware bi-leveloptimization model for multi-job scheduling problems undercloud computingrdquo Soft Computing vol 20 no 1 pp 303ndash3172016

[85] R Yanggratoke F Wuhib and R Stadler ldquoGossip-basedresource allocation for green computing in large cloudsrdquo inProceedings of the 2011 7th International Conference onNetwork and Service Management pp 1ndash9 IEEE ParisFrance October 2011

[86] A Paya and D C Marinescu ldquoEnergy-aware load balancingand application scaling for the cloud ecosystemrdquo IEEETransactions on Cloud Computing vol 5 no 1 pp 15ndash272015

[87] V D Justafort R Beaubrun and S Pierre ldquoA hybrid ap-proach for optimizing carbon footprint in intercloud envi-ronmentrdquo IEEE Transactions on Services Computing vol 12no 2 pp 186ndash198 2016

[88] R Panwar and B Mallick ldquoLoad balancing in cloud com-puting using dynamic load management algorithmrdquo inProceedings of the 2015 International Conference on GreenComputing and Internet of Aings (ICGCIoT) pp 773ndash778IEEE Noida India October 2015

[89] C-T Yang K-C Wang H-Y Cheng C-T Kuo andW C C Chu ldquoGreen power management with dynamicresource allocation for cloud virtual machinesrdquo in Proceedingsof the 2011 IEEE International Conference on High Perfor-mance Computing and Communications pp 726ndash733 IEEEAlberta Canada September 2011

[90] C-T Yang H-Y Cheng and K-L Huang ldquoA dynamicresource allocation model for virtual machine management

Complexity 15

on cloudrdquo in Proceedings of the International Conference onGrid and Distributed Computing pp 581ndash590 SpringerGangneug Korea December 2011

[91] X Xu F Shucun C Qing et al ldquoDynamic resource allocationfor load balancing in fog environmentrdquo Wireless Commu-nications amp Mobile Computing vol 2018 no 2 Article ID6421607 15 pages 2018

[92] J Oueis E C Strinati and S Barbarossa ldquoe fog balancingload distribution for small cell cloud computingrdquo in Pro-ceedings of the 2015 IEEE 81st Vehicular Technology Confer-ence (VTC Spring) pp 1ndash6 IEEE Glasgow UK May 2015

[93] K Wang Y Wang Y Sun S Guo and J Wu ldquoGreen in-dustrial internet of things architecture an energy-efficientperspectiverdquo IEEE Communications Magazine vol 54 no 12pp 48ndash54 2016

16 Complexity

Li et al [69] proposed communication framework in 5Gand studied the problem of allocating power andchannels So the signal data in the channel can beavailable and the total energy efficiency can be maximumFormula (9) [69] shows how to calculate the energy ef-ficiency of a system

U 1113944K

k11113944

M

i1EES

iK + 1113944K

K11113944

N

i1EEA

jk (9)

where EESiK and EEA

jk respectively denote the energy ef-ficiency of sensor S and actuator A on channels e sets ofsensors actuators and channels are respectively repre-sented as s S1 S2 SM1113864 1113865 A A1 A2 AN1113864 1113865 andC C1 C2 CK1113864 1113865

Liu et al [70] studied the problem of allocating resourcesefficiently on IoT that supplies wireless power In thismethod users are first grouped into accessible channelsAnd then power distribution of users grouped in the samechannel is studied to improve throughput of the networkis method can allocate finite resources to a large group ofusers Ejaz and Ibnkahla [71] proposed the resource allo-cation framework with several bands under cognitive 5GIoT In the highly dynamic environment of the IoTmultiband method can manage resources more flexiblyand reduce more energy consumption In addition areconstruction approach with several levels is proposed toallocate resources reasonably for applications with dif-ferent needs of QoS Colistra et al [72] proposed aprotocol which is distributed and optimal to allocateresources in heterogeneous IoT Because this protocol hasexcellent adaptability when changing topology of net-work it can distribute resources evenly among nodes Jianet al [73] proposed a multilevel allocating resources al-gorithm for IoT communication using advanced tech-nology e algorithm uses hierarchical structure and hasfast data processing rate and very low latency in a satu-rated or not saturated environment

Zheng and Liu [74] proposed a new algorithm to allocatebandwidth dynamically for controlling remote computers inthe IoT is method can reduce the error of signal re-construction under the same bandwidth and make thebandwidth allocation of IoT more reasonable Gai and Qiu[75] used reinforcement learning mechanisms to allocateresources to achieve high Quality of Experience ismethod can effectively solve the resource allocation prob-lems caused by the mismatch of service quality and complexservice providing condition in the IoT

Table 4 shows a summary of the above works eproblem in Table 4 represents the realization of dynamic andflexible allocation of resources e resources here canrepresent channels bandwidth and power

23 Scientific Workflow Execution Implementing scientificworkflows especially in heterogeneous environments canreduce resource waste and reduce energy costs Scientificworkflow can be obtained by reasonably allocating resourcesand dynamically deploying VMs

231 Cloud Computing Xu et al [76] proposed a resourceallocation method called EnRealan to solve the problem ofenergy consumption e dynamic deployment of VMs isgenerally adopted to execute scientific workflows Bousselmiet al [77] proposed a scheduling method based on energyperception for executing scientific workflows in cloudcomputing At first an algorithm of splitting workflows forenergy minimization is presented which can achieve a highparallelism without huge energy consumption en aheuristic algorithm used to optimize cat swarm is proposedfor the created partitions e algorithm can minimize thetotal consumption of energy and the execution time ofworkflows Sonia et al [78] proposed a workflow schedulingmethod with several objects and hybrid particle swarmoptimization algorithm In addition a method for dy-namically scaling voltage and frequency is proposed emethod canmake the processors work at any voltage level soas to minimize the energy consumption in the process ofworkflow scheduling Both Bousselmi et al [77] and Soniaet al [78] use scheduling method to achieve scientificworkflows and study the problem of energy consumptione difference is that Bousselmi et al [77] focuses on in-tensive computing tasks while Sonia et al [78] focuses onworkflow scheduling on heterogeneous computing systems

Cao [79] established a scheduling algorithm of scientificworkflows with an objective of energy savingis algorithmcan enable service providers to gain high profits and reduceusersrsquo overhead at the same time Li et al [80] proposed ascheduling algorithm based on cloud computing which canminimize cost of performing workflows within a specifiedtime In addition the rented VM was modified to save costfurther Khaleel and Zhu [81] proposed a scheduling algo-rithm and took scientific workflows as a model to make fulluse of cloud resources and save energy Shi et al [82]designed a flexible resource allocation and job schedulingmechanism to implement scientific workflows Because thismechanism can implement scientific workflows withinprescribed budgets and deadlines so it can work better thanother mechanisms

Table 5 shows a summary of the abovementioned workse problems in Table 5 are derived from the imple-mentation of the scientific workflow VM deployment refersto the rational allocation of VMs Workflow scheduling refersto reducing the scheduling energy and time In addition itrefers to the scheduling of the workflow on heterogeneoussystems Cost reduction refers to reducing the cost of workflowexecution Effective implementation is about scientific work-flow execution within a specified budget and time

24 Server Optimization Server optimization is also a goodway to save energy e goal of optimizing the server can beachieved by uninstalling unnecessary servers or consoli-dating servers as well as by reasonably scheduling tasksUnlike QoS optimization server optimization aims to op-timize the number of used servers improves the energyefficiency of servers and consolidates servers However QoSoptimization studies how tomake users get better experienceand meet their needs

10 Complexity

241 Cloud Computing Ge et al [83] proposed a game-theoretic method and transformed the problem of mini-mizing energy into a congestion game All mobile devices inthis method are participants in the game e methodchooses a server to unload the computation tasks to optimizeQoS levels and save energy which can optimize the systemand save energy Wang et al [84] proposed a MapReduce-based multitask scheduling algorithm to achieve the ob-jective of energy saving is model is a two-layer modelwhich considers the impact of server performance changeson energy consumption and the limitation of networkbandwidth In addition a local search operator is designedbased on which a two-layer genetic algorithm is proposede algorithm can schedule tens of thousands of tasks incloud and achieve large-scale optimization Yanggratoke

et al [85] proposed a general generic gossip protocol aimingat allocating resources in cloud environment An instanti-ation of this protocol was developed to enable server con-solidation to allocate resources to more servers to meetchanging load patterns

25 LoadBalancing Load balancing can help save energy bymanaging the number of servers and allocating resources

251 Cloud Computing Paya and Marinescu [86] intro-duced an operation model that balances cloud computingload and expands applications to save energy e principleof this model is to define an operating system e systemshould make servers run in the system as many as possible

Table 5 Work summary of scientific workflow execution in cloud computing

Problems Solutions Literatures Advantages

VM deployment A resource allocation method namedEnRealan [76] Performs scientific workflows based on energy perception

across cloud platforms

Workflowscheduling

A scheduling method based on energyperception [77]

Achieves a high parallelism without huge energyconsumption and minimizes the total consumption of

energy and execution time of workflows

A workflow scheduling method with severalobjects and hybrid particle swarm

optimization algorithm[78]

Makes the processors work at any voltage level minimizesthe energy consumption in the process of workflowscheduling and studies the scheduling problem of

workflows on heterogeneous systemsA scheduling algorithm based on various

applications [79] Enables service providers to gain high profits and reducesuser overhead at the same time

Cost reduction A scheduling algorithm based on energyperception [80 81] Minimizes the cost of performing workflows while

meeting the time constraintEffectiveimplementation

A flexible resource allocation and jobscheduling mechanism [82] Implements scientific workflows within prescribed

budgets and deadlines

Table 4 Work summary of resource allocation and management in IoT

Problems Solutions Literatures Advantages

Resourcemanagement

A distributed cloud network framework [66]Replaces the centralized architecture with a distributed cloud

architecture solves the defects of the centralized cloudarchitecture and brings people better experience

A MILP model [68] Assigns the requirement of servicesrsquo resource to heterogeneousnetwork equipment interface

A framework for communication used in5G [69]

Transforms the resource allocation problem into a power andchannel allocation problem minimizes the total energy

consumption and improves QoS levelsA low complexity channel allocation

algorithm [70] Improves throughput of the network and allocates finiteresources to a large group of users

A resource allocation framework withseveral bands under cognitive 5G IoT [71] Manages resources more flexibly and reduces energy

consumption than common single-band approach

A protocol which is distributed andoptimal to allocate resources [72]

Has excellent adaptability when changing topology of networkand dynamically manages resources in the heterogeneous IoT

environmentA multilevel allocating resourcesalgorithm for IoT communication [73] Has fast data processing rate and very low latency in both

saturated and nonsaturated environment

A new algorithm to allocate bandwidthdynamically [74]

Reduces the error of signal reconstruction under the samebandwidth and makes the bandwidth allocation of IoT more

reasonable

A reinforcement learning mechanism [75]Effectively solves the resource allocation problems caused by themismatch of service quality and complex service providing

condition in the IoT

Complexity 11

When no tasks are being performed the system should adjustservers to sleep thus energy consumption can be reducedJustafort et al [87] mainly studied the problem of workloaddistribution across cloud computing environment and pro-posed a method to solve the problem of the VM layout So thefootprint of carbon can be effectively reduced Panwar andMallick [88] proposed an algorithm to dynamically manage theload and effectively distribute the total incoming requestsbetween VMs rough efficient and uniform utilization ofresources this algorithm can achieve uniform distribution ofload between servers Yang et al [89] proposed a powermanagement mechanism to balance the load e system canmonitor VMs and dynamically allocate the resources Yanget al [90] proposed an optimization system to better allocateresources dynamically which can balance the load of VMsrunning on multiple physical machines Under this systemVMs can be migrated automatically to adjust high and lowloads without interrupting services Yang et al [89 90] manageVMs to achieve load balancing ey allocate resources dy-namically to migrate VMs which can balance workloads ondifferent physical machines e difference is that Yanget al[89] integrate a dynamic resource allocation approachwithOpenNebulaWhile Yang et al [90] focuse on avoiding serviceoutages during VM migration

Table 6 shows a summary of the abovementioned workse problems in Table 6 are from the load balancing problemServer management is about the control of the number ofservers running in the system Workload management is therational allocation of workload or tasks VMmanagement refersto configuringVMresources andmigratingVMs to adjust loads

252 Fog Computing Xu et al [91] proposed a methodcalled ldquoDRAMrdquo to dynamically allocate resources in fogcomputing environment which can avoid both too highand too low loads e method first analyzes the loadbalance of different kinds of computing nodes And then itdesigns a fog resource allocation method to achieve loadbalance which allocates resources statically and migratesservices dynamically Oueis et al [92] studied the loadbalance problem in fog computing A custom fog clus-tering algorithm is proposed to solve the problem In theproblem several users need to offload computations andall of their demands need to be handled by local com-puting cluster

253 IoT Wang et al [93] established architecture of theenergy saving targeted system in industrial IoT In additionin order to predict sleep intervals they developed a sleepscheduling and a wake protocol which provide a better wayfor energy saving

3 Conclusion

is paper did a comprehensive study of QoS optimi-zation and energy saving in cloud computing edgecomputing fog computing and IoT models We sum-marized five main problems and analyzed their solutionsproposed by existing works By conducting this surveywe aim to help readers have a deeper understanding onthe concepts of different computing models and thetechniques of QoS optimization and energy saving inthese models

e investigated papers focus on issues about ensuringQoS and reducing SLA violations and resource manage-ment In the case of QoS assurance and SLA violationsreduction the main solution of QoS assurance is efficientVM management is solution can meet customersrsquo re-quirements through reasonable scheduling and integrationof VMs Most of resource management techniques are re-alized by reasonable scheduling of resources which canreduce the waste of VMs servers and traffic

Disclosure

is manuscript is an extension of A Survey of QoS Opti-mization and Energy Saving in Cloud Edge and IoT in e9th EAI International Conference on Cloud Computing

Conflicts of Interest

e authors declare that there are no conflicts of interestregarding the publication of this paper

Acknowledgments

is work was supported by the National Natural ScienceFoundation of China (Grant no 61702274) Natural ScienceFoundation of Jiangsu Province (Grant no BK20170958)and PAPD

Table 6 Work summary of load balancing in cloud computing

Problems Solutions Literatures AdvantagesServermanagement

An operation model that can balance cloudcomputing load and expand application [86] Saves energy by managing the number of servers running in

the system

Workloadmanagement

A hybrid approach [87] Reduces the footprint of carbon and allocates workload acrosscloud computing

An algorithm to dynamically manage theload [88] Manages the load evens the load distribution between

servers and allocates tasks between VMs

VMmanagement

A green power administration mechanism [89] Monitors and jointly allocates VM resources

An optimization system [90]Migrates VMs to adjust high and low loads without

interrupting services and balances the load of VMs runningon multiple physical machines

12 Complexity

References

[1] Q Zhang L Cheng and R Boutaba ldquoCloud computingstate-of-the-art and research challengesrdquo Journal of InternetServices and Applications vol 1 no 1 pp 7ndash18 2010

[2] W Shi J Cao Q Zhang Y Li and L Xu ldquoEdge computingvision and challengesrdquo IEEE Internet of Aings Journal vol 3no 5 pp 637ndash646 2016

[3] M Iorga L Feldman R Barton M J Martin N S Gorenand C Mahmoudi ldquoFog computing conceptual modelrdquo TechRep Recommendations of the National Institute of Standardsand Technology Gaithersburg MD USA 2018

[4] Nebbiolo ldquoFog vs edge computingrdquo Tech Rep NebbioloTechnologies Inc Milpitas CA USA 2018

[5] C T Do N H Tran C Pham M G R Alam J H Son andC S Hong ldquoA proximal algorithm for joint resource allo-cation andminimizing carbon footprint in geo-distributed fogcomputingrdquo in Proceedings of the 2015 International Con-ference on Information Networking (ICOIN) pp 324ndash329IEEE Siem Reap Cambodia January 2015

[6] S Vashi J Ram J Modi S Verma and C Prakash ldquoInternetof things (IoT) a vision architectural elements and securityissuesrdquo in Proceedings of the 2017 International Conference onI-SMAC (IoT in Social Mobile Analytics and Cloud)(I-SMAC) pp 492ndash496 IEEE Coimbatore India February2017

[7] M Mazzucco D Dyachuk and R Deters ldquoMaximizing cloudprovidersrsquo revenues via energy aware allocation policiesrdquo inProceedings of the 2010 IEEE 3rd International Conference onCloud Computing pp 131ndash138 IEEE Miami FL USA July2010

[8] Q He R Zhou X Zhang et al ldquoKeyword search for buildingservice-based systemsrdquo IEEE Transactions on Software En-gineering vol 43 no 7 pp 658ndash674 2016

[9] L Sun H Dong O K Hussain F K Hussain and A X LiuldquoA framework of cloud service selection with criteria inter-actionsrdquo Future Generation Computer Systems vol 94pp 749ndash764 2019

[10] M Mazzucco and D Dyachuk ldquoOptimizing cloud providersrevenues via energy efficient server allocationrdquo SustainableComputing Informatics and Systems vol 2 no 1 pp 1ndash122012

[11] Q He J Han F Chen et al ldquoQos-aware service selection forcustomisable multi-tenant service-based systems maturityand approachesrdquo in Proceedings of the 2015 IEEE 8th Inter-national Conference on Cloud Computing pp 237ndash244 IEEENew York NY USA July 2015

[12] L Sun J Ma H Wang Y Zhang and J Yong ldquoCloud servicedescription model an extension of usdl for cloud servicesrdquoIEEE Transactions on Services Computing vol 11 no 2pp 354ndash368 2015

[13] Y Wang Q He D Ye and Y Yang ldquoFormulating criticality-based cost-effective fault tolerance strategies for multi-tenantservice-based systemsrdquo IEEE Transactions on Software En-gineering vol 44 no 3 pp 291ndash307 2017

[14] S Mustafa K Bilal S U R Malik and S A Madani ldquoSla-aware energy efficient resource management for cloud en-vironmentsrdquo IEEE Access vol 6 pp 15004ndash15020 2018

[15] J Bi H Yuan M Tie and W Tan ldquoSla-based optimisation ofvirtualised resource for multi-tier web applications in clouddata centresrdquo Enterprise Information Systems vol 9 no 7pp 743ndash767 2015

[16] S Singh I Chana and R Buyya ldquoStar sla-aware autonomicmanagement of cloud resourcesrdquo IEEE Transactions on CloudComputing p 1 2017

[17] A Beloglazov and R Buyya ldquoEnergy efficient resourcemanagement in virtualized cloud data centersrdquo in Proceedingsof the 2010 10th IEEEACM International Conference onCluster Cloud and Grid Computing pp 826ndash831 IEEEComputer Society Melbourne Australia May 2010

[18] M Guazzone C Anglano and M Canonico ldquoEnergy-effi-cient resource management for cloud computing infra-structuresrdquo in Proceedings of the 2011 IEEE AirdInternational Conference on Cloud Computing Technology andScience pp 424ndash431 IEEE Athens Greece November 2011

[19] Y Sun J White and S Eade ldquoA model-based system toautomate cloud resource allocation and optimizationrdquo inProceedings of the International Conference on Model DrivenEngineering Languages and Systems pp 18ndash34 SpringerValencia Spain October 2014

[20] G Siddesh and K Srinivasa ldquoSla-driven dynamic resourceallocation on cloudsrdquo in Proceedings of the InternationalConference on Advanced Computing Networking and Securitypp 9ndash18 Springer Surathkal India December 2011

[21] S K Garg S K Gopalaiyengar and R Buyya ldquoSla-basedresource provisioning for heterogeneous workloads in avirtualized cloud datacenterrdquo in Proceedings of the Interna-tional Conference on Algorithms and Architectures for ParallelProcessing pp 371ndash384 Springer Melbourne AustraliaOctober 2011

[22] J Bi Z Zhu and H Yuan ldquoSla-aware dynamic resourceprovisioning for profit maximization in shared cloud datacentersrdquo in Proceedings of the International Conference onHigh Performance Networking Computing and Communi-cation Systems pp 366ndash372 Springer Singapore May 2011

[23] L Qi Y Chen Y Yuan S Fu X Zhang and X Xu ldquoA Qos-aware virtual machine scheduling method for energy con-servation in cloud-based cyber-physical systemsrdquo WorldWide Web vol 4 no 3 pp 1ndash23 2019

[24] A Beloglazov J Abawajy and R Buyya ldquoEnergy-aware re-source allocation heuristics for efficient management of datacenters for cloud computingrdquo Future Generation ComputerSystems vol 28 no 5 pp 755ndash768 2012

[25] L Qi R Wang C Hu S Li Q He and X Xu ldquoTime-awaredistributed service recommendation with privacy-preserva-tionrdquo Information Sciences vol 480 pp 354ndash364 2019

[26] Z Zhai B Cheng Y Tian J Chen L Zhao and M Niu ldquoAdata-driven service creation approach for end-usersrdquo IEEEAccess vol 4 pp 9923ndash9940 2016

[27] L Gu D Zeng S Guo A Barnawi and Y Xiang ldquoCostefficient resource management in fog computing supportedmedical cyber-physical systemrdquo IEEE Transactions onEmerging Topics in Computing vol 5 no 1 pp 108ndash119 2015

[28] L Ni J Zhang C Jiang C Yan and K Yu ldquoResource al-location strategy in fog computing based on priced timed petrinetsrdquo Ieee Internet of Aings Journal vol 4 no 5pp 1216ndash1228 2017

[29] L Wei T Yang F C Delicato et al ldquoOn enabling sustainableedge computing with renewable energy resourcesrdquo IEEECommunications Magazine vol 56 no 5 pp 94ndash101 2018

[30] P Lai Q He G Cui et al ldquoEdge user allocation with dynamicquality of servicerdquo in Proceedings of the International Con-ference on Service-Oriented Computing pp 86ndash101 SpringerToulouse France October 2019

[31] X Xu Q Liu X Zhang J Zhang L Qi and W Dou ldquoAblockchain-powered crowdsourcing method with privacy

Complexity 13

preservation in mobile environmentrdquo IEEE Transactions onComputational Social Systems vol 6 no 6 pp 1407ndash1419 2019

[32] O Rolik E Zharikov and S Telenyk ldquoMicrocloud-basedarchitecture of management system for IoT infrastructuresrdquoin Proceedings of the 2016 Aird International Scientific-Practical Conference Problems of Infocommunications Scienceand Technology (PIC SampT) pp 149ndash151 IEEE KharkivUkaraine October 2016

[33] W He S Guo Y Liang R Ma X Qiu and L Shi ldquoQos-awareand resource-efficient dynamic slicingmechanism for internetof thingsrdquo Computers Materials amp Continua vol 61 no 3pp 1345ndash1364 2019

[34] J Yao and N Ansari ldquoQos-aware fog resource provisioningand mobile device power control in IoT networksrdquo IEEETransactions on Network and Service Management vol 16no 1 pp 167ndash175 2018

[35] W Wang Y Jiang and W Wu ldquoMultiagent-based resourceallocation for energy minimization in cloud computing sys-temsrdquo IEEE Transactions on Systems Man and CyberneticsSystems vol 47 no 2 pp 1ndash16 2016

[36] K Krishnajyothi ldquoParallel data processing for effective dy-namic resource allocation in the cloudrdquo International Journalof Computer Applications vol 70 no 22 pp 1ndash4 2013

[37] M M Hassan B Song M S Hossain and A Alamri ldquoEf-ficient resource scheduling for big data processing in cloudplatformrdquo in Proceedings of the International Conference onInternet and Distributed Computing Systems pp 51ndash63Springer Calabria Italy September 2014

[38] C-M Wu R-S Chang and H-Y Chan ldquoA green energy-efficient scheduling algorithm using the dvfs technique forcloud datacentersrdquo Future Generation Computer Systemsvol 37 no 7 pp 141ndash147 2014

[39] Z Wang and X Su ldquoDynamically hierarchical resource-al-location algorithm in cloud computing environmentrdquo AeJournal of Supercomputing vol 71 no 7 pp 2748ndash2766 2015

[40] W-Y Lin G-Y Lin and H-Y Wei ldquoDynamic auctionmechanism for cloud resource allocationrdquo in Proceedings ofthe 2010 10th IEEEACM International Conference on ClusterCloud and Grid Computing pp 591-592 IEEE ComputerSociety Victoria Australia May 2010

[41] Y O Yazir C Matthews R Farahbod et al ldquoDynamic re-source allocation in computing clouds using distributedmultiple criteria decision analysisrdquo in Proceedings of the 2010IEEE 3rd International Conference on Cloud Computingpp 91ndash98 IEEE Washington DC USA July 2010

[42] W Lin J Z Wang C Liang and D Qi ldquoA threshold-baseddynamic resource allocation scheme for cloud computingrdquoProcedia Engineering vol 23 no 5 pp 695ndash703 2011

[43] X Xu S Fu L Qi et al ldquoAn IoT-oriented data placementmethod with privacy preservation in cloud environmentrdquoJournal of Network and Computer Applications vol 124pp 148ndash157 2018

[44] J Bokyun P Md Jalil L Daeho and S Doug Young ldquoEf-ficient computation offloading in mobile cloud computing forvideo streaming over 5 grdquo Computers Materials amp Continuavol 61 no 2 pp 439ndash463 2019

[45] H Sarbazi-Azad and A Y Zomaya ldquoEnergy-efficient resourceutilization in cloud computingrdquo in Large Scale Network-Centric Distributed Systems pp 377ndash408 Wiley-IEEE PressHoboken NJ USA 2014

[46] C-H Hsu K D Slagter S-C Chen and Y-C ChungldquoOptimizing energy consumption with task consolidation incloudsrdquo Information Sciences vol 258 no 3 pp 452ndash4622014

[47] C-H Hsu S-C Chen C-C Lee et al ldquoEnergy-aware taskconsolidation technique for cloud computingrdquo in Proceedingsof the 2011 IEEE Aird International Conference on CloudComputing Technology and Science pp 115ndash121 IEEE Ath-ens Greece 2011

[48] S K Panda and P K Jana ldquoAn efficient task consolidationalgorithm for cloud computing systemsrdquo in Proceedings of theInternational Conference on Distributed Computing and In-ternet Technology pp 61ndash74 Springer Bhubaneswar IndiaJanuary 2016

[49] L Yin J Luo and H Luo ldquoTasks scheduling and resourceallocation in fog computing based on containers for smartmanufacturingrdquo IEEE Transactions on Industrial Informaticsvol 14 no 10 pp 4712ndash4721 2018

[50] M Aazam and E-N Huh ldquoDynamic resource provisioningthrough fog micro datacenterrdquo in Proceedings of the 2015IEEE International Conference on Pervasive Computing andCommunication Workshops (PerCom Workshops) pp 105ndash110 IEEE St Louis MO USA March 2015

[51] B Jia H Hu Y Zeng T Xu and Y Yang ldquoDouble-matchingresource allocation strategy in fog computing networks basedon cost efficiencyrdquo Journal of Communications and Networksvol 20 no 3 pp 237ndash246 2018

[52] H Zhang Y Xiao S Bu D Niyato F R Yu and Z HanldquoComputing resource allocation in three-tier IoT fog net-works a joint optimization approach combining stackelberggame and matchingrdquo IEEE Internet of Aings Journal vol 4no 5 pp 1204ndash1215 2017

[53] J Tan T-H Chang and T Q Quelc ldquoMinimum energyresource allocation in fog radio access network with fronthauland latency constraintsrdquo in Proceedings of the 2018 IEEE 19thInternational Workshop on Signal Processing Advances inWireless Communications (SPAWC) pp 1ndash5 IEEE KalamataGreece June 2018

[54] D R de Vasconcelos R M de Castro Andrade andJ N de Souza ldquoSmart shadowndashan autonomous availabilitycomputation resource allocation platform for internet ofthings in the fog computing environmentrdquo in Proceedings of the2015 International Conference on Distributed Computing inSensor Systems pp 216-217 IEEE Fortaleza Brazil June 2015

[55] M Aazam M St-Hilaire C-H Lung and I LambadarisldquoPre-fog IoT trace based probabilistic resource estimation atfogrdquo in Proceedings of the 2016 13th IEEE Annual ConsumerCommunications amp Networking Conference (CCNC) pp 12ndash17 IEEE Las Vegas NV USA January 2016

[56] N D Tung L L Bao and B Vijay ldquoPrice-based resourceallocation for edge computing a market equilibrium ap-proachrdquo IEEE Transactions on Cloud Computing p 1 2018

[57] X Xu C He Z Xu L Qi S Wan and M Z A BhuiyanldquoJoint optimization of offloading utility and privacy for edgecomputing enabled IoTrdquo IEEE Internet of Aings Journal2019

[58] X Xu Y Chen X Zhang Q Liu X Liu and L Qi ldquoAblockchain-based computation offloading method for edgecomputing in 5G networksrdquo Software Practice and Experi-ence 2019

[59] X Xu Y Li T Huang et al ldquoAn energy-aware computationoffloading method for smart edge computing in wirelessmetropolitan area networksrdquo Journal of Network and Com-puter Applications vol 133 pp 75ndash85 2019

[60] X Xu Y Xue L Qi et al ldquoAn edge computing-enabledcomputation offloading method with privacy preservation forinternet of connected vehiclesrdquo Future Generation ComputerSystems vol 96 pp 89ndash100 2019

14 Complexity

[61] G Yeting L Fang X Nong and C Zhengguo ldquoTask-basedresource allocation bid in edge computing micro datacenterrdquoComputers Materials amp Continua vol 61 no 2 pp 777ndash7922019

[62] X Chen L Jiao W Li and X Fu ldquoEfficient multi-usercomputation offloading for mobile-edge cloud computingrdquoIEEEACM Transactions on Networking vol 24 no 5pp 2795ndash2808 2016

[63] B Gao L He X Lu C Chang K Li and K Li ldquoDevelopingenergy-aware task allocation schemes in cloud-assisted mo-bile workflowsrdquo in Proceedings of the 2015 IEEE InternationalConference on Computer and Information Technology Ubiq-uitous Computing and Communications Dependable Auto-nomic and Secure Computing Pervasive Intelligence andComputing pp 1266ndash1273 IEEE Liverpool UK October2015

[64] X Xu X Zhang H Gao Y Xue L Qi andW Dou ldquoBecomeblockchain-enabled computation offloading for IoT in mobileedge computingrdquo IEEE Transactions on Industrial Infor-matics 2019

[65] W Yifei W Zhaoying G Da and Y F Richard ldquoDeepq-learning based computation offloading strategy for mobileedge computingrdquo Computers Materials amp Continua vol 59no 1 pp 89ndash104 2019

[66] M Barcelo A Correa J Llorca A M Tulino J L Vicarioand A Morell ldquoIoT-cloud service optimization in nextgeneration smart environmentsrdquo IEEE Journal on SelectedAreas in Communications vol 34 no 12 pp 4077ndash40902016

[67] B Cheng M Wang S Zhao Z Zhai D Zhu and J ChenldquoSituation-aware dynamic service coordination in an IoTenvironmentrdquo IEEEACM Transactions On Networkingvol 25 no 4 pp 2082ndash2095 2017

[68] V Angelakis I Avgouleas N Pappas and D Yuan ldquoFlexibleallocation of heterogeneous resources to services on an IoTdevicerdquo in Proceedings of the 2015 IEEE Conference onComputer Communications Workshops (INFOCOMWKSHPS) pp 99-100 IEEE Hong Kong China April 2015

[69] S Li Q Ni Y Sun G Min and S Al-Rubaye ldquoEnergy-ef-ficient resource allocation for industrial cyber-physical IoTsystems in 5g erardquo IEEE Transactions on Industrial Infor-matics vol 14 no 6 pp 2618ndash2628 2018

[70] X Liu Z Qin Y Gao and J A McCann ldquoResource allocationin wireless powered IoT networksrdquo IEEE Internet of AingsJournal vol 6 no 3 pp 4935ndash4945 2019

[71] W Ejaz and M Ibnkahla ldquoMulti-band spectrum sensing andresource allocation for IoT in cognitive 5g networksrdquo IEEEInternet of Aings Journal vol 5 no 1 pp 150ndash163 2017

[72] G Colistra V Pilloni and L Atzori ldquoTask allocation in groupof nodes in the IoT a consensus approachrdquo in Proceedings ofthe 2014 IEEE International Conference on Communications(ICC) pp 3848ndash3853 IEEE Sydney Australia June 2014

[73] J Li Q Sun and G Fan ldquoResource allocation for multiclassservice in IoTuplink communicationsrdquo inProceedings of the 20163rd International Conference on Systems and Informatics (ICSAI)pp 777ndash781 IEEE Shanghai China November 2016

[74] L I Zheng and K H Liu ldquoDynamic bandwidth resourceallocation algorithm in internet of things and its applicationrdquoComputer Engineering vol 38 no 17 pp 16ndash19 2012

[75] K Gai and M Qiu ldquoOptimal resource allocation using re-inforcement learning for IoT content-centric servicesrdquo Ap-plied Soft Computing vol 70 pp 12ndash21 2018

[76] X Xu W Dou X Zhang and J Chen ldquoEnreal an energy-aware resource allocation method for scientific workflow

executions in cloud environmentrdquo IEEE Transactions onCloud Computing vol 4 no 2 pp 166ndash179 2016

[77] K Bousselmi Z Brahmi and M M Gammoudi ldquoEnergyefficient partitioning and scheduling approach for scientificworkflows in the cloudrdquo in Proceedings of the 2016 IEEEInternational Conference on Services Computing (SCC)pp 146ndash154 IEEE San Francisco CA USA June 2016

[78] Y Sonia C Rachid K Hubert and G Bertrand ldquoMulti-objective approach for energy-aware workflow scheduling incloud computing environmentsrdquo Ae Scientific World Jour-nal vol 2013 Article ID 350934 13 pages 2013

[79] F Cao Efficient Scientific Workflow Scheduling in CloudEnvironment Southern Illinois University Carbondale ILUSA 2014

[80] Z Li J Ge H Hu W Song H Hu and B Luo ldquoCost andenergy aware scheduling algorithm for scientific workflowswith deadline constraint in cloudsrdquo IEEE Transactions onServices Computing vol 11 no 4 pp 713ndash726 2015

[81] M Khaleel and M M Zhu ldquoEnergy-aware job managementapproaches for workflow in cloudrdquo in Proceedings of the 2015IEEE International Conference on Cluster Computingpp 506-507 IEEE Chicago IL USA September 2015

[82] J Shi J Luo F Dong and J Zhang ldquoA budget and deadlineaware scientific workflow resource provisioning and sched-uling mechanism for cloudrdquo in Proceedings of the 2014 IEEE18th International Conference on Computer Supported Co-operative Work in Design (CSCWD) pp 672ndash677 IEEEHsinchu Taiwan January 2014

[83] Y Ge Y Zhang Q Qiu and Y-H Lu ldquoA game theoreticresource allocation for overall energy minimization in mobilecloud computing systemrdquo in Proceedings of the 2012ACMIEEE International Symposium on Low Power Electronicsand Design pp 279ndash284 ACM Redondo Beach CA USAAugust 2012

[84] X Wang Y Wang and C Yue ldquoAn energy-aware bi-leveloptimization model for multi-job scheduling problems undercloud computingrdquo Soft Computing vol 20 no 1 pp 303ndash3172016

[85] R Yanggratoke F Wuhib and R Stadler ldquoGossip-basedresource allocation for green computing in large cloudsrdquo inProceedings of the 2011 7th International Conference onNetwork and Service Management pp 1ndash9 IEEE ParisFrance October 2011

[86] A Paya and D C Marinescu ldquoEnergy-aware load balancingand application scaling for the cloud ecosystemrdquo IEEETransactions on Cloud Computing vol 5 no 1 pp 15ndash272015

[87] V D Justafort R Beaubrun and S Pierre ldquoA hybrid ap-proach for optimizing carbon footprint in intercloud envi-ronmentrdquo IEEE Transactions on Services Computing vol 12no 2 pp 186ndash198 2016

[88] R Panwar and B Mallick ldquoLoad balancing in cloud com-puting using dynamic load management algorithmrdquo inProceedings of the 2015 International Conference on GreenComputing and Internet of Aings (ICGCIoT) pp 773ndash778IEEE Noida India October 2015

[89] C-T Yang K-C Wang H-Y Cheng C-T Kuo andW C C Chu ldquoGreen power management with dynamicresource allocation for cloud virtual machinesrdquo in Proceedingsof the 2011 IEEE International Conference on High Perfor-mance Computing and Communications pp 726ndash733 IEEEAlberta Canada September 2011

[90] C-T Yang H-Y Cheng and K-L Huang ldquoA dynamicresource allocation model for virtual machine management

Complexity 15

on cloudrdquo in Proceedings of the International Conference onGrid and Distributed Computing pp 581ndash590 SpringerGangneug Korea December 2011

[91] X Xu F Shucun C Qing et al ldquoDynamic resource allocationfor load balancing in fog environmentrdquo Wireless Commu-nications amp Mobile Computing vol 2018 no 2 Article ID6421607 15 pages 2018

[92] J Oueis E C Strinati and S Barbarossa ldquoe fog balancingload distribution for small cell cloud computingrdquo in Pro-ceedings of the 2015 IEEE 81st Vehicular Technology Confer-ence (VTC Spring) pp 1ndash6 IEEE Glasgow UK May 2015

[93] K Wang Y Wang Y Sun S Guo and J Wu ldquoGreen in-dustrial internet of things architecture an energy-efficientperspectiverdquo IEEE Communications Magazine vol 54 no 12pp 48ndash54 2016

16 Complexity

241 Cloud Computing Ge et al [83] proposed a game-theoretic method and transformed the problem of mini-mizing energy into a congestion game All mobile devices inthis method are participants in the game e methodchooses a server to unload the computation tasks to optimizeQoS levels and save energy which can optimize the systemand save energy Wang et al [84] proposed a MapReduce-based multitask scheduling algorithm to achieve the ob-jective of energy saving is model is a two-layer modelwhich considers the impact of server performance changeson energy consumption and the limitation of networkbandwidth In addition a local search operator is designedbased on which a two-layer genetic algorithm is proposede algorithm can schedule tens of thousands of tasks incloud and achieve large-scale optimization Yanggratoke

et al [85] proposed a general generic gossip protocol aimingat allocating resources in cloud environment An instanti-ation of this protocol was developed to enable server con-solidation to allocate resources to more servers to meetchanging load patterns

25 LoadBalancing Load balancing can help save energy bymanaging the number of servers and allocating resources

251 Cloud Computing Paya and Marinescu [86] intro-duced an operation model that balances cloud computingload and expands applications to save energy e principleof this model is to define an operating system e systemshould make servers run in the system as many as possible

Table 5 Work summary of scientific workflow execution in cloud computing

Problems Solutions Literatures Advantages

VM deployment A resource allocation method namedEnRealan [76] Performs scientific workflows based on energy perception

across cloud platforms

Workflowscheduling

A scheduling method based on energyperception [77]

Achieves a high parallelism without huge energyconsumption and minimizes the total consumption of

energy and execution time of workflows

A workflow scheduling method with severalobjects and hybrid particle swarm

optimization algorithm[78]

Makes the processors work at any voltage level minimizesthe energy consumption in the process of workflowscheduling and studies the scheduling problem of

workflows on heterogeneous systemsA scheduling algorithm based on various

applications [79] Enables service providers to gain high profits and reducesuser overhead at the same time

Cost reduction A scheduling algorithm based on energyperception [80 81] Minimizes the cost of performing workflows while

meeting the time constraintEffectiveimplementation

A flexible resource allocation and jobscheduling mechanism [82] Implements scientific workflows within prescribed

budgets and deadlines

Table 4 Work summary of resource allocation and management in IoT

Problems Solutions Literatures Advantages

Resourcemanagement

A distributed cloud network framework [66]Replaces the centralized architecture with a distributed cloud

architecture solves the defects of the centralized cloudarchitecture and brings people better experience

A MILP model [68] Assigns the requirement of servicesrsquo resource to heterogeneousnetwork equipment interface

A framework for communication used in5G [69]

Transforms the resource allocation problem into a power andchannel allocation problem minimizes the total energy

consumption and improves QoS levelsA low complexity channel allocation

algorithm [70] Improves throughput of the network and allocates finiteresources to a large group of users

A resource allocation framework withseveral bands under cognitive 5G IoT [71] Manages resources more flexibly and reduces energy

consumption than common single-band approach

A protocol which is distributed andoptimal to allocate resources [72]

Has excellent adaptability when changing topology of networkand dynamically manages resources in the heterogeneous IoT

environmentA multilevel allocating resourcesalgorithm for IoT communication [73] Has fast data processing rate and very low latency in both

saturated and nonsaturated environment

A new algorithm to allocate bandwidthdynamically [74]

Reduces the error of signal reconstruction under the samebandwidth and makes the bandwidth allocation of IoT more

reasonable

A reinforcement learning mechanism [75]Effectively solves the resource allocation problems caused by themismatch of service quality and complex service providing

condition in the IoT

Complexity 11

When no tasks are being performed the system should adjustservers to sleep thus energy consumption can be reducedJustafort et al [87] mainly studied the problem of workloaddistribution across cloud computing environment and pro-posed a method to solve the problem of the VM layout So thefootprint of carbon can be effectively reduced Panwar andMallick [88] proposed an algorithm to dynamically manage theload and effectively distribute the total incoming requestsbetween VMs rough efficient and uniform utilization ofresources this algorithm can achieve uniform distribution ofload between servers Yang et al [89] proposed a powermanagement mechanism to balance the load e system canmonitor VMs and dynamically allocate the resources Yanget al [90] proposed an optimization system to better allocateresources dynamically which can balance the load of VMsrunning on multiple physical machines Under this systemVMs can be migrated automatically to adjust high and lowloads without interrupting services Yang et al [89 90] manageVMs to achieve load balancing ey allocate resources dy-namically to migrate VMs which can balance workloads ondifferent physical machines e difference is that Yanget al[89] integrate a dynamic resource allocation approachwithOpenNebulaWhile Yang et al [90] focuse on avoiding serviceoutages during VM migration

Table 6 shows a summary of the abovementioned workse problems in Table 6 are from the load balancing problemServer management is about the control of the number ofservers running in the system Workload management is therational allocation of workload or tasks VMmanagement refersto configuringVMresources andmigratingVMs to adjust loads

252 Fog Computing Xu et al [91] proposed a methodcalled ldquoDRAMrdquo to dynamically allocate resources in fogcomputing environment which can avoid both too highand too low loads e method first analyzes the loadbalance of different kinds of computing nodes And then itdesigns a fog resource allocation method to achieve loadbalance which allocates resources statically and migratesservices dynamically Oueis et al [92] studied the loadbalance problem in fog computing A custom fog clus-tering algorithm is proposed to solve the problem In theproblem several users need to offload computations andall of their demands need to be handled by local com-puting cluster

253 IoT Wang et al [93] established architecture of theenergy saving targeted system in industrial IoT In additionin order to predict sleep intervals they developed a sleepscheduling and a wake protocol which provide a better wayfor energy saving

3 Conclusion

is paper did a comprehensive study of QoS optimi-zation and energy saving in cloud computing edgecomputing fog computing and IoT models We sum-marized five main problems and analyzed their solutionsproposed by existing works By conducting this surveywe aim to help readers have a deeper understanding onthe concepts of different computing models and thetechniques of QoS optimization and energy saving inthese models

e investigated papers focus on issues about ensuringQoS and reducing SLA violations and resource manage-ment In the case of QoS assurance and SLA violationsreduction the main solution of QoS assurance is efficientVM management is solution can meet customersrsquo re-quirements through reasonable scheduling and integrationof VMs Most of resource management techniques are re-alized by reasonable scheduling of resources which canreduce the waste of VMs servers and traffic

Disclosure

is manuscript is an extension of A Survey of QoS Opti-mization and Energy Saving in Cloud Edge and IoT in e9th EAI International Conference on Cloud Computing

Conflicts of Interest

e authors declare that there are no conflicts of interestregarding the publication of this paper

Acknowledgments

is work was supported by the National Natural ScienceFoundation of China (Grant no 61702274) Natural ScienceFoundation of Jiangsu Province (Grant no BK20170958)and PAPD

Table 6 Work summary of load balancing in cloud computing

Problems Solutions Literatures AdvantagesServermanagement

An operation model that can balance cloudcomputing load and expand application [86] Saves energy by managing the number of servers running in

the system

Workloadmanagement

A hybrid approach [87] Reduces the footprint of carbon and allocates workload acrosscloud computing

An algorithm to dynamically manage theload [88] Manages the load evens the load distribution between

servers and allocates tasks between VMs

VMmanagement

A green power administration mechanism [89] Monitors and jointly allocates VM resources

An optimization system [90]Migrates VMs to adjust high and low loads without

interrupting services and balances the load of VMs runningon multiple physical machines

12 Complexity

References

[1] Q Zhang L Cheng and R Boutaba ldquoCloud computingstate-of-the-art and research challengesrdquo Journal of InternetServices and Applications vol 1 no 1 pp 7ndash18 2010

[2] W Shi J Cao Q Zhang Y Li and L Xu ldquoEdge computingvision and challengesrdquo IEEE Internet of Aings Journal vol 3no 5 pp 637ndash646 2016

[3] M Iorga L Feldman R Barton M J Martin N S Gorenand C Mahmoudi ldquoFog computing conceptual modelrdquo TechRep Recommendations of the National Institute of Standardsand Technology Gaithersburg MD USA 2018

[4] Nebbiolo ldquoFog vs edge computingrdquo Tech Rep NebbioloTechnologies Inc Milpitas CA USA 2018

[5] C T Do N H Tran C Pham M G R Alam J H Son andC S Hong ldquoA proximal algorithm for joint resource allo-cation andminimizing carbon footprint in geo-distributed fogcomputingrdquo in Proceedings of the 2015 International Con-ference on Information Networking (ICOIN) pp 324ndash329IEEE Siem Reap Cambodia January 2015

[6] S Vashi J Ram J Modi S Verma and C Prakash ldquoInternetof things (IoT) a vision architectural elements and securityissuesrdquo in Proceedings of the 2017 International Conference onI-SMAC (IoT in Social Mobile Analytics and Cloud)(I-SMAC) pp 492ndash496 IEEE Coimbatore India February2017

[7] M Mazzucco D Dyachuk and R Deters ldquoMaximizing cloudprovidersrsquo revenues via energy aware allocation policiesrdquo inProceedings of the 2010 IEEE 3rd International Conference onCloud Computing pp 131ndash138 IEEE Miami FL USA July2010

[8] Q He R Zhou X Zhang et al ldquoKeyword search for buildingservice-based systemsrdquo IEEE Transactions on Software En-gineering vol 43 no 7 pp 658ndash674 2016

[9] L Sun H Dong O K Hussain F K Hussain and A X LiuldquoA framework of cloud service selection with criteria inter-actionsrdquo Future Generation Computer Systems vol 94pp 749ndash764 2019

[10] M Mazzucco and D Dyachuk ldquoOptimizing cloud providersrevenues via energy efficient server allocationrdquo SustainableComputing Informatics and Systems vol 2 no 1 pp 1ndash122012

[11] Q He J Han F Chen et al ldquoQos-aware service selection forcustomisable multi-tenant service-based systems maturityand approachesrdquo in Proceedings of the 2015 IEEE 8th Inter-national Conference on Cloud Computing pp 237ndash244 IEEENew York NY USA July 2015

[12] L Sun J Ma H Wang Y Zhang and J Yong ldquoCloud servicedescription model an extension of usdl for cloud servicesrdquoIEEE Transactions on Services Computing vol 11 no 2pp 354ndash368 2015

[13] Y Wang Q He D Ye and Y Yang ldquoFormulating criticality-based cost-effective fault tolerance strategies for multi-tenantservice-based systemsrdquo IEEE Transactions on Software En-gineering vol 44 no 3 pp 291ndash307 2017

[14] S Mustafa K Bilal S U R Malik and S A Madani ldquoSla-aware energy efficient resource management for cloud en-vironmentsrdquo IEEE Access vol 6 pp 15004ndash15020 2018

[15] J Bi H Yuan M Tie and W Tan ldquoSla-based optimisation ofvirtualised resource for multi-tier web applications in clouddata centresrdquo Enterprise Information Systems vol 9 no 7pp 743ndash767 2015

[16] S Singh I Chana and R Buyya ldquoStar sla-aware autonomicmanagement of cloud resourcesrdquo IEEE Transactions on CloudComputing p 1 2017

[17] A Beloglazov and R Buyya ldquoEnergy efficient resourcemanagement in virtualized cloud data centersrdquo in Proceedingsof the 2010 10th IEEEACM International Conference onCluster Cloud and Grid Computing pp 826ndash831 IEEEComputer Society Melbourne Australia May 2010

[18] M Guazzone C Anglano and M Canonico ldquoEnergy-effi-cient resource management for cloud computing infra-structuresrdquo in Proceedings of the 2011 IEEE AirdInternational Conference on Cloud Computing Technology andScience pp 424ndash431 IEEE Athens Greece November 2011

[19] Y Sun J White and S Eade ldquoA model-based system toautomate cloud resource allocation and optimizationrdquo inProceedings of the International Conference on Model DrivenEngineering Languages and Systems pp 18ndash34 SpringerValencia Spain October 2014

[20] G Siddesh and K Srinivasa ldquoSla-driven dynamic resourceallocation on cloudsrdquo in Proceedings of the InternationalConference on Advanced Computing Networking and Securitypp 9ndash18 Springer Surathkal India December 2011

[21] S K Garg S K Gopalaiyengar and R Buyya ldquoSla-basedresource provisioning for heterogeneous workloads in avirtualized cloud datacenterrdquo in Proceedings of the Interna-tional Conference on Algorithms and Architectures for ParallelProcessing pp 371ndash384 Springer Melbourne AustraliaOctober 2011

[22] J Bi Z Zhu and H Yuan ldquoSla-aware dynamic resourceprovisioning for profit maximization in shared cloud datacentersrdquo in Proceedings of the International Conference onHigh Performance Networking Computing and Communi-cation Systems pp 366ndash372 Springer Singapore May 2011

[23] L Qi Y Chen Y Yuan S Fu X Zhang and X Xu ldquoA Qos-aware virtual machine scheduling method for energy con-servation in cloud-based cyber-physical systemsrdquo WorldWide Web vol 4 no 3 pp 1ndash23 2019

[24] A Beloglazov J Abawajy and R Buyya ldquoEnergy-aware re-source allocation heuristics for efficient management of datacenters for cloud computingrdquo Future Generation ComputerSystems vol 28 no 5 pp 755ndash768 2012

[25] L Qi R Wang C Hu S Li Q He and X Xu ldquoTime-awaredistributed service recommendation with privacy-preserva-tionrdquo Information Sciences vol 480 pp 354ndash364 2019

[26] Z Zhai B Cheng Y Tian J Chen L Zhao and M Niu ldquoAdata-driven service creation approach for end-usersrdquo IEEEAccess vol 4 pp 9923ndash9940 2016

[27] L Gu D Zeng S Guo A Barnawi and Y Xiang ldquoCostefficient resource management in fog computing supportedmedical cyber-physical systemrdquo IEEE Transactions onEmerging Topics in Computing vol 5 no 1 pp 108ndash119 2015

[28] L Ni J Zhang C Jiang C Yan and K Yu ldquoResource al-location strategy in fog computing based on priced timed petrinetsrdquo Ieee Internet of Aings Journal vol 4 no 5pp 1216ndash1228 2017

[29] L Wei T Yang F C Delicato et al ldquoOn enabling sustainableedge computing with renewable energy resourcesrdquo IEEECommunications Magazine vol 56 no 5 pp 94ndash101 2018

[30] P Lai Q He G Cui et al ldquoEdge user allocation with dynamicquality of servicerdquo in Proceedings of the International Con-ference on Service-Oriented Computing pp 86ndash101 SpringerToulouse France October 2019

[31] X Xu Q Liu X Zhang J Zhang L Qi and W Dou ldquoAblockchain-powered crowdsourcing method with privacy

Complexity 13

preservation in mobile environmentrdquo IEEE Transactions onComputational Social Systems vol 6 no 6 pp 1407ndash1419 2019

[32] O Rolik E Zharikov and S Telenyk ldquoMicrocloud-basedarchitecture of management system for IoT infrastructuresrdquoin Proceedings of the 2016 Aird International Scientific-Practical Conference Problems of Infocommunications Scienceand Technology (PIC SampT) pp 149ndash151 IEEE KharkivUkaraine October 2016

[33] W He S Guo Y Liang R Ma X Qiu and L Shi ldquoQos-awareand resource-efficient dynamic slicingmechanism for internetof thingsrdquo Computers Materials amp Continua vol 61 no 3pp 1345ndash1364 2019

[34] J Yao and N Ansari ldquoQos-aware fog resource provisioningand mobile device power control in IoT networksrdquo IEEETransactions on Network and Service Management vol 16no 1 pp 167ndash175 2018

[35] W Wang Y Jiang and W Wu ldquoMultiagent-based resourceallocation for energy minimization in cloud computing sys-temsrdquo IEEE Transactions on Systems Man and CyberneticsSystems vol 47 no 2 pp 1ndash16 2016

[36] K Krishnajyothi ldquoParallel data processing for effective dy-namic resource allocation in the cloudrdquo International Journalof Computer Applications vol 70 no 22 pp 1ndash4 2013

[37] M M Hassan B Song M S Hossain and A Alamri ldquoEf-ficient resource scheduling for big data processing in cloudplatformrdquo in Proceedings of the International Conference onInternet and Distributed Computing Systems pp 51ndash63Springer Calabria Italy September 2014

[38] C-M Wu R-S Chang and H-Y Chan ldquoA green energy-efficient scheduling algorithm using the dvfs technique forcloud datacentersrdquo Future Generation Computer Systemsvol 37 no 7 pp 141ndash147 2014

[39] Z Wang and X Su ldquoDynamically hierarchical resource-al-location algorithm in cloud computing environmentrdquo AeJournal of Supercomputing vol 71 no 7 pp 2748ndash2766 2015

[40] W-Y Lin G-Y Lin and H-Y Wei ldquoDynamic auctionmechanism for cloud resource allocationrdquo in Proceedings ofthe 2010 10th IEEEACM International Conference on ClusterCloud and Grid Computing pp 591-592 IEEE ComputerSociety Victoria Australia May 2010

[41] Y O Yazir C Matthews R Farahbod et al ldquoDynamic re-source allocation in computing clouds using distributedmultiple criteria decision analysisrdquo in Proceedings of the 2010IEEE 3rd International Conference on Cloud Computingpp 91ndash98 IEEE Washington DC USA July 2010

[42] W Lin J Z Wang C Liang and D Qi ldquoA threshold-baseddynamic resource allocation scheme for cloud computingrdquoProcedia Engineering vol 23 no 5 pp 695ndash703 2011

[43] X Xu S Fu L Qi et al ldquoAn IoT-oriented data placementmethod with privacy preservation in cloud environmentrdquoJournal of Network and Computer Applications vol 124pp 148ndash157 2018

[44] J Bokyun P Md Jalil L Daeho and S Doug Young ldquoEf-ficient computation offloading in mobile cloud computing forvideo streaming over 5 grdquo Computers Materials amp Continuavol 61 no 2 pp 439ndash463 2019

[45] H Sarbazi-Azad and A Y Zomaya ldquoEnergy-efficient resourceutilization in cloud computingrdquo in Large Scale Network-Centric Distributed Systems pp 377ndash408 Wiley-IEEE PressHoboken NJ USA 2014

[46] C-H Hsu K D Slagter S-C Chen and Y-C ChungldquoOptimizing energy consumption with task consolidation incloudsrdquo Information Sciences vol 258 no 3 pp 452ndash4622014

[47] C-H Hsu S-C Chen C-C Lee et al ldquoEnergy-aware taskconsolidation technique for cloud computingrdquo in Proceedingsof the 2011 IEEE Aird International Conference on CloudComputing Technology and Science pp 115ndash121 IEEE Ath-ens Greece 2011

[48] S K Panda and P K Jana ldquoAn efficient task consolidationalgorithm for cloud computing systemsrdquo in Proceedings of theInternational Conference on Distributed Computing and In-ternet Technology pp 61ndash74 Springer Bhubaneswar IndiaJanuary 2016

[49] L Yin J Luo and H Luo ldquoTasks scheduling and resourceallocation in fog computing based on containers for smartmanufacturingrdquo IEEE Transactions on Industrial Informaticsvol 14 no 10 pp 4712ndash4721 2018

[50] M Aazam and E-N Huh ldquoDynamic resource provisioningthrough fog micro datacenterrdquo in Proceedings of the 2015IEEE International Conference on Pervasive Computing andCommunication Workshops (PerCom Workshops) pp 105ndash110 IEEE St Louis MO USA March 2015

[51] B Jia H Hu Y Zeng T Xu and Y Yang ldquoDouble-matchingresource allocation strategy in fog computing networks basedon cost efficiencyrdquo Journal of Communications and Networksvol 20 no 3 pp 237ndash246 2018

[52] H Zhang Y Xiao S Bu D Niyato F R Yu and Z HanldquoComputing resource allocation in three-tier IoT fog net-works a joint optimization approach combining stackelberggame and matchingrdquo IEEE Internet of Aings Journal vol 4no 5 pp 1204ndash1215 2017

[53] J Tan T-H Chang and T Q Quelc ldquoMinimum energyresource allocation in fog radio access network with fronthauland latency constraintsrdquo in Proceedings of the 2018 IEEE 19thInternational Workshop on Signal Processing Advances inWireless Communications (SPAWC) pp 1ndash5 IEEE KalamataGreece June 2018

[54] D R de Vasconcelos R M de Castro Andrade andJ N de Souza ldquoSmart shadowndashan autonomous availabilitycomputation resource allocation platform for internet ofthings in the fog computing environmentrdquo in Proceedings of the2015 International Conference on Distributed Computing inSensor Systems pp 216-217 IEEE Fortaleza Brazil June 2015

[55] M Aazam M St-Hilaire C-H Lung and I LambadarisldquoPre-fog IoT trace based probabilistic resource estimation atfogrdquo in Proceedings of the 2016 13th IEEE Annual ConsumerCommunications amp Networking Conference (CCNC) pp 12ndash17 IEEE Las Vegas NV USA January 2016

[56] N D Tung L L Bao and B Vijay ldquoPrice-based resourceallocation for edge computing a market equilibrium ap-proachrdquo IEEE Transactions on Cloud Computing p 1 2018

[57] X Xu C He Z Xu L Qi S Wan and M Z A BhuiyanldquoJoint optimization of offloading utility and privacy for edgecomputing enabled IoTrdquo IEEE Internet of Aings Journal2019

[58] X Xu Y Chen X Zhang Q Liu X Liu and L Qi ldquoAblockchain-based computation offloading method for edgecomputing in 5G networksrdquo Software Practice and Experi-ence 2019

[59] X Xu Y Li T Huang et al ldquoAn energy-aware computationoffloading method for smart edge computing in wirelessmetropolitan area networksrdquo Journal of Network and Com-puter Applications vol 133 pp 75ndash85 2019

[60] X Xu Y Xue L Qi et al ldquoAn edge computing-enabledcomputation offloading method with privacy preservation forinternet of connected vehiclesrdquo Future Generation ComputerSystems vol 96 pp 89ndash100 2019

14 Complexity

[61] G Yeting L Fang X Nong and C Zhengguo ldquoTask-basedresource allocation bid in edge computing micro datacenterrdquoComputers Materials amp Continua vol 61 no 2 pp 777ndash7922019

[62] X Chen L Jiao W Li and X Fu ldquoEfficient multi-usercomputation offloading for mobile-edge cloud computingrdquoIEEEACM Transactions on Networking vol 24 no 5pp 2795ndash2808 2016

[63] B Gao L He X Lu C Chang K Li and K Li ldquoDevelopingenergy-aware task allocation schemes in cloud-assisted mo-bile workflowsrdquo in Proceedings of the 2015 IEEE InternationalConference on Computer and Information Technology Ubiq-uitous Computing and Communications Dependable Auto-nomic and Secure Computing Pervasive Intelligence andComputing pp 1266ndash1273 IEEE Liverpool UK October2015

[64] X Xu X Zhang H Gao Y Xue L Qi andW Dou ldquoBecomeblockchain-enabled computation offloading for IoT in mobileedge computingrdquo IEEE Transactions on Industrial Infor-matics 2019

[65] W Yifei W Zhaoying G Da and Y F Richard ldquoDeepq-learning based computation offloading strategy for mobileedge computingrdquo Computers Materials amp Continua vol 59no 1 pp 89ndash104 2019

[66] M Barcelo A Correa J Llorca A M Tulino J L Vicarioand A Morell ldquoIoT-cloud service optimization in nextgeneration smart environmentsrdquo IEEE Journal on SelectedAreas in Communications vol 34 no 12 pp 4077ndash40902016

[67] B Cheng M Wang S Zhao Z Zhai D Zhu and J ChenldquoSituation-aware dynamic service coordination in an IoTenvironmentrdquo IEEEACM Transactions On Networkingvol 25 no 4 pp 2082ndash2095 2017

[68] V Angelakis I Avgouleas N Pappas and D Yuan ldquoFlexibleallocation of heterogeneous resources to services on an IoTdevicerdquo in Proceedings of the 2015 IEEE Conference onComputer Communications Workshops (INFOCOMWKSHPS) pp 99-100 IEEE Hong Kong China April 2015

[69] S Li Q Ni Y Sun G Min and S Al-Rubaye ldquoEnergy-ef-ficient resource allocation for industrial cyber-physical IoTsystems in 5g erardquo IEEE Transactions on Industrial Infor-matics vol 14 no 6 pp 2618ndash2628 2018

[70] X Liu Z Qin Y Gao and J A McCann ldquoResource allocationin wireless powered IoT networksrdquo IEEE Internet of AingsJournal vol 6 no 3 pp 4935ndash4945 2019

[71] W Ejaz and M Ibnkahla ldquoMulti-band spectrum sensing andresource allocation for IoT in cognitive 5g networksrdquo IEEEInternet of Aings Journal vol 5 no 1 pp 150ndash163 2017

[72] G Colistra V Pilloni and L Atzori ldquoTask allocation in groupof nodes in the IoT a consensus approachrdquo in Proceedings ofthe 2014 IEEE International Conference on Communications(ICC) pp 3848ndash3853 IEEE Sydney Australia June 2014

[73] J Li Q Sun and G Fan ldquoResource allocation for multiclassservice in IoTuplink communicationsrdquo inProceedings of the 20163rd International Conference on Systems and Informatics (ICSAI)pp 777ndash781 IEEE Shanghai China November 2016

[74] L I Zheng and K H Liu ldquoDynamic bandwidth resourceallocation algorithm in internet of things and its applicationrdquoComputer Engineering vol 38 no 17 pp 16ndash19 2012

[75] K Gai and M Qiu ldquoOptimal resource allocation using re-inforcement learning for IoT content-centric servicesrdquo Ap-plied Soft Computing vol 70 pp 12ndash21 2018

[76] X Xu W Dou X Zhang and J Chen ldquoEnreal an energy-aware resource allocation method for scientific workflow

executions in cloud environmentrdquo IEEE Transactions onCloud Computing vol 4 no 2 pp 166ndash179 2016

[77] K Bousselmi Z Brahmi and M M Gammoudi ldquoEnergyefficient partitioning and scheduling approach for scientificworkflows in the cloudrdquo in Proceedings of the 2016 IEEEInternational Conference on Services Computing (SCC)pp 146ndash154 IEEE San Francisco CA USA June 2016

[78] Y Sonia C Rachid K Hubert and G Bertrand ldquoMulti-objective approach for energy-aware workflow scheduling incloud computing environmentsrdquo Ae Scientific World Jour-nal vol 2013 Article ID 350934 13 pages 2013

[79] F Cao Efficient Scientific Workflow Scheduling in CloudEnvironment Southern Illinois University Carbondale ILUSA 2014

[80] Z Li J Ge H Hu W Song H Hu and B Luo ldquoCost andenergy aware scheduling algorithm for scientific workflowswith deadline constraint in cloudsrdquo IEEE Transactions onServices Computing vol 11 no 4 pp 713ndash726 2015

[81] M Khaleel and M M Zhu ldquoEnergy-aware job managementapproaches for workflow in cloudrdquo in Proceedings of the 2015IEEE International Conference on Cluster Computingpp 506-507 IEEE Chicago IL USA September 2015

[82] J Shi J Luo F Dong and J Zhang ldquoA budget and deadlineaware scientific workflow resource provisioning and sched-uling mechanism for cloudrdquo in Proceedings of the 2014 IEEE18th International Conference on Computer Supported Co-operative Work in Design (CSCWD) pp 672ndash677 IEEEHsinchu Taiwan January 2014

[83] Y Ge Y Zhang Q Qiu and Y-H Lu ldquoA game theoreticresource allocation for overall energy minimization in mobilecloud computing systemrdquo in Proceedings of the 2012ACMIEEE International Symposium on Low Power Electronicsand Design pp 279ndash284 ACM Redondo Beach CA USAAugust 2012

[84] X Wang Y Wang and C Yue ldquoAn energy-aware bi-leveloptimization model for multi-job scheduling problems undercloud computingrdquo Soft Computing vol 20 no 1 pp 303ndash3172016

[85] R Yanggratoke F Wuhib and R Stadler ldquoGossip-basedresource allocation for green computing in large cloudsrdquo inProceedings of the 2011 7th International Conference onNetwork and Service Management pp 1ndash9 IEEE ParisFrance October 2011

[86] A Paya and D C Marinescu ldquoEnergy-aware load balancingand application scaling for the cloud ecosystemrdquo IEEETransactions on Cloud Computing vol 5 no 1 pp 15ndash272015

[87] V D Justafort R Beaubrun and S Pierre ldquoA hybrid ap-proach for optimizing carbon footprint in intercloud envi-ronmentrdquo IEEE Transactions on Services Computing vol 12no 2 pp 186ndash198 2016

[88] R Panwar and B Mallick ldquoLoad balancing in cloud com-puting using dynamic load management algorithmrdquo inProceedings of the 2015 International Conference on GreenComputing and Internet of Aings (ICGCIoT) pp 773ndash778IEEE Noida India October 2015

[89] C-T Yang K-C Wang H-Y Cheng C-T Kuo andW C C Chu ldquoGreen power management with dynamicresource allocation for cloud virtual machinesrdquo in Proceedingsof the 2011 IEEE International Conference on High Perfor-mance Computing and Communications pp 726ndash733 IEEEAlberta Canada September 2011

[90] C-T Yang H-Y Cheng and K-L Huang ldquoA dynamicresource allocation model for virtual machine management

Complexity 15

on cloudrdquo in Proceedings of the International Conference onGrid and Distributed Computing pp 581ndash590 SpringerGangneug Korea December 2011

[91] X Xu F Shucun C Qing et al ldquoDynamic resource allocationfor load balancing in fog environmentrdquo Wireless Commu-nications amp Mobile Computing vol 2018 no 2 Article ID6421607 15 pages 2018

[92] J Oueis E C Strinati and S Barbarossa ldquoe fog balancingload distribution for small cell cloud computingrdquo in Pro-ceedings of the 2015 IEEE 81st Vehicular Technology Confer-ence (VTC Spring) pp 1ndash6 IEEE Glasgow UK May 2015

[93] K Wang Y Wang Y Sun S Guo and J Wu ldquoGreen in-dustrial internet of things architecture an energy-efficientperspectiverdquo IEEE Communications Magazine vol 54 no 12pp 48ndash54 2016

16 Complexity

When no tasks are being performed the system should adjustservers to sleep thus energy consumption can be reducedJustafort et al [87] mainly studied the problem of workloaddistribution across cloud computing environment and pro-posed a method to solve the problem of the VM layout So thefootprint of carbon can be effectively reduced Panwar andMallick [88] proposed an algorithm to dynamically manage theload and effectively distribute the total incoming requestsbetween VMs rough efficient and uniform utilization ofresources this algorithm can achieve uniform distribution ofload between servers Yang et al [89] proposed a powermanagement mechanism to balance the load e system canmonitor VMs and dynamically allocate the resources Yanget al [90] proposed an optimization system to better allocateresources dynamically which can balance the load of VMsrunning on multiple physical machines Under this systemVMs can be migrated automatically to adjust high and lowloads without interrupting services Yang et al [89 90] manageVMs to achieve load balancing ey allocate resources dy-namically to migrate VMs which can balance workloads ondifferent physical machines e difference is that Yanget al[89] integrate a dynamic resource allocation approachwithOpenNebulaWhile Yang et al [90] focuse on avoiding serviceoutages during VM migration

Table 6 shows a summary of the abovementioned workse problems in Table 6 are from the load balancing problemServer management is about the control of the number ofservers running in the system Workload management is therational allocation of workload or tasks VMmanagement refersto configuringVMresources andmigratingVMs to adjust loads

252 Fog Computing Xu et al [91] proposed a methodcalled ldquoDRAMrdquo to dynamically allocate resources in fogcomputing environment which can avoid both too highand too low loads e method first analyzes the loadbalance of different kinds of computing nodes And then itdesigns a fog resource allocation method to achieve loadbalance which allocates resources statically and migratesservices dynamically Oueis et al [92] studied the loadbalance problem in fog computing A custom fog clus-tering algorithm is proposed to solve the problem In theproblem several users need to offload computations andall of their demands need to be handled by local com-puting cluster

253 IoT Wang et al [93] established architecture of theenergy saving targeted system in industrial IoT In additionin order to predict sleep intervals they developed a sleepscheduling and a wake protocol which provide a better wayfor energy saving

3 Conclusion

is paper did a comprehensive study of QoS optimi-zation and energy saving in cloud computing edgecomputing fog computing and IoT models We sum-marized five main problems and analyzed their solutionsproposed by existing works By conducting this surveywe aim to help readers have a deeper understanding onthe concepts of different computing models and thetechniques of QoS optimization and energy saving inthese models

e investigated papers focus on issues about ensuringQoS and reducing SLA violations and resource manage-ment In the case of QoS assurance and SLA violationsreduction the main solution of QoS assurance is efficientVM management is solution can meet customersrsquo re-quirements through reasonable scheduling and integrationof VMs Most of resource management techniques are re-alized by reasonable scheduling of resources which canreduce the waste of VMs servers and traffic

Disclosure

is manuscript is an extension of A Survey of QoS Opti-mization and Energy Saving in Cloud Edge and IoT in e9th EAI International Conference on Cloud Computing

Conflicts of Interest

e authors declare that there are no conflicts of interestregarding the publication of this paper

Acknowledgments

is work was supported by the National Natural ScienceFoundation of China (Grant no 61702274) Natural ScienceFoundation of Jiangsu Province (Grant no BK20170958)and PAPD

Table 6 Work summary of load balancing in cloud computing

Problems Solutions Literatures AdvantagesServermanagement

An operation model that can balance cloudcomputing load and expand application [86] Saves energy by managing the number of servers running in

the system

Workloadmanagement

A hybrid approach [87] Reduces the footprint of carbon and allocates workload acrosscloud computing

An algorithm to dynamically manage theload [88] Manages the load evens the load distribution between

servers and allocates tasks between VMs

VMmanagement

A green power administration mechanism [89] Monitors and jointly allocates VM resources

An optimization system [90]Migrates VMs to adjust high and low loads without

interrupting services and balances the load of VMs runningon multiple physical machines

12 Complexity

References

[1] Q Zhang L Cheng and R Boutaba ldquoCloud computingstate-of-the-art and research challengesrdquo Journal of InternetServices and Applications vol 1 no 1 pp 7ndash18 2010

[2] W Shi J Cao Q Zhang Y Li and L Xu ldquoEdge computingvision and challengesrdquo IEEE Internet of Aings Journal vol 3no 5 pp 637ndash646 2016

[3] M Iorga L Feldman R Barton M J Martin N S Gorenand C Mahmoudi ldquoFog computing conceptual modelrdquo TechRep Recommendations of the National Institute of Standardsand Technology Gaithersburg MD USA 2018

[4] Nebbiolo ldquoFog vs edge computingrdquo Tech Rep NebbioloTechnologies Inc Milpitas CA USA 2018

[5] C T Do N H Tran C Pham M G R Alam J H Son andC S Hong ldquoA proximal algorithm for joint resource allo-cation andminimizing carbon footprint in geo-distributed fogcomputingrdquo in Proceedings of the 2015 International Con-ference on Information Networking (ICOIN) pp 324ndash329IEEE Siem Reap Cambodia January 2015

[6] S Vashi J Ram J Modi S Verma and C Prakash ldquoInternetof things (IoT) a vision architectural elements and securityissuesrdquo in Proceedings of the 2017 International Conference onI-SMAC (IoT in Social Mobile Analytics and Cloud)(I-SMAC) pp 492ndash496 IEEE Coimbatore India February2017

[7] M Mazzucco D Dyachuk and R Deters ldquoMaximizing cloudprovidersrsquo revenues via energy aware allocation policiesrdquo inProceedings of the 2010 IEEE 3rd International Conference onCloud Computing pp 131ndash138 IEEE Miami FL USA July2010

[8] Q He R Zhou X Zhang et al ldquoKeyword search for buildingservice-based systemsrdquo IEEE Transactions on Software En-gineering vol 43 no 7 pp 658ndash674 2016

[9] L Sun H Dong O K Hussain F K Hussain and A X LiuldquoA framework of cloud service selection with criteria inter-actionsrdquo Future Generation Computer Systems vol 94pp 749ndash764 2019

[10] M Mazzucco and D Dyachuk ldquoOptimizing cloud providersrevenues via energy efficient server allocationrdquo SustainableComputing Informatics and Systems vol 2 no 1 pp 1ndash122012

[11] Q He J Han F Chen et al ldquoQos-aware service selection forcustomisable multi-tenant service-based systems maturityand approachesrdquo in Proceedings of the 2015 IEEE 8th Inter-national Conference on Cloud Computing pp 237ndash244 IEEENew York NY USA July 2015

[12] L Sun J Ma H Wang Y Zhang and J Yong ldquoCloud servicedescription model an extension of usdl for cloud servicesrdquoIEEE Transactions on Services Computing vol 11 no 2pp 354ndash368 2015

[13] Y Wang Q He D Ye and Y Yang ldquoFormulating criticality-based cost-effective fault tolerance strategies for multi-tenantservice-based systemsrdquo IEEE Transactions on Software En-gineering vol 44 no 3 pp 291ndash307 2017

[14] S Mustafa K Bilal S U R Malik and S A Madani ldquoSla-aware energy efficient resource management for cloud en-vironmentsrdquo IEEE Access vol 6 pp 15004ndash15020 2018

[15] J Bi H Yuan M Tie and W Tan ldquoSla-based optimisation ofvirtualised resource for multi-tier web applications in clouddata centresrdquo Enterprise Information Systems vol 9 no 7pp 743ndash767 2015

[16] S Singh I Chana and R Buyya ldquoStar sla-aware autonomicmanagement of cloud resourcesrdquo IEEE Transactions on CloudComputing p 1 2017

[17] A Beloglazov and R Buyya ldquoEnergy efficient resourcemanagement in virtualized cloud data centersrdquo in Proceedingsof the 2010 10th IEEEACM International Conference onCluster Cloud and Grid Computing pp 826ndash831 IEEEComputer Society Melbourne Australia May 2010

[18] M Guazzone C Anglano and M Canonico ldquoEnergy-effi-cient resource management for cloud computing infra-structuresrdquo in Proceedings of the 2011 IEEE AirdInternational Conference on Cloud Computing Technology andScience pp 424ndash431 IEEE Athens Greece November 2011

[19] Y Sun J White and S Eade ldquoA model-based system toautomate cloud resource allocation and optimizationrdquo inProceedings of the International Conference on Model DrivenEngineering Languages and Systems pp 18ndash34 SpringerValencia Spain October 2014

[20] G Siddesh and K Srinivasa ldquoSla-driven dynamic resourceallocation on cloudsrdquo in Proceedings of the InternationalConference on Advanced Computing Networking and Securitypp 9ndash18 Springer Surathkal India December 2011

[21] S K Garg S K Gopalaiyengar and R Buyya ldquoSla-basedresource provisioning for heterogeneous workloads in avirtualized cloud datacenterrdquo in Proceedings of the Interna-tional Conference on Algorithms and Architectures for ParallelProcessing pp 371ndash384 Springer Melbourne AustraliaOctober 2011

[22] J Bi Z Zhu and H Yuan ldquoSla-aware dynamic resourceprovisioning for profit maximization in shared cloud datacentersrdquo in Proceedings of the International Conference onHigh Performance Networking Computing and Communi-cation Systems pp 366ndash372 Springer Singapore May 2011

[23] L Qi Y Chen Y Yuan S Fu X Zhang and X Xu ldquoA Qos-aware virtual machine scheduling method for energy con-servation in cloud-based cyber-physical systemsrdquo WorldWide Web vol 4 no 3 pp 1ndash23 2019

[24] A Beloglazov J Abawajy and R Buyya ldquoEnergy-aware re-source allocation heuristics for efficient management of datacenters for cloud computingrdquo Future Generation ComputerSystems vol 28 no 5 pp 755ndash768 2012

[25] L Qi R Wang C Hu S Li Q He and X Xu ldquoTime-awaredistributed service recommendation with privacy-preserva-tionrdquo Information Sciences vol 480 pp 354ndash364 2019

[26] Z Zhai B Cheng Y Tian J Chen L Zhao and M Niu ldquoAdata-driven service creation approach for end-usersrdquo IEEEAccess vol 4 pp 9923ndash9940 2016

[27] L Gu D Zeng S Guo A Barnawi and Y Xiang ldquoCostefficient resource management in fog computing supportedmedical cyber-physical systemrdquo IEEE Transactions onEmerging Topics in Computing vol 5 no 1 pp 108ndash119 2015

[28] L Ni J Zhang C Jiang C Yan and K Yu ldquoResource al-location strategy in fog computing based on priced timed petrinetsrdquo Ieee Internet of Aings Journal vol 4 no 5pp 1216ndash1228 2017

[29] L Wei T Yang F C Delicato et al ldquoOn enabling sustainableedge computing with renewable energy resourcesrdquo IEEECommunications Magazine vol 56 no 5 pp 94ndash101 2018

[30] P Lai Q He G Cui et al ldquoEdge user allocation with dynamicquality of servicerdquo in Proceedings of the International Con-ference on Service-Oriented Computing pp 86ndash101 SpringerToulouse France October 2019

[31] X Xu Q Liu X Zhang J Zhang L Qi and W Dou ldquoAblockchain-powered crowdsourcing method with privacy

Complexity 13

preservation in mobile environmentrdquo IEEE Transactions onComputational Social Systems vol 6 no 6 pp 1407ndash1419 2019

[32] O Rolik E Zharikov and S Telenyk ldquoMicrocloud-basedarchitecture of management system for IoT infrastructuresrdquoin Proceedings of the 2016 Aird International Scientific-Practical Conference Problems of Infocommunications Scienceand Technology (PIC SampT) pp 149ndash151 IEEE KharkivUkaraine October 2016

[33] W He S Guo Y Liang R Ma X Qiu and L Shi ldquoQos-awareand resource-efficient dynamic slicingmechanism for internetof thingsrdquo Computers Materials amp Continua vol 61 no 3pp 1345ndash1364 2019

[34] J Yao and N Ansari ldquoQos-aware fog resource provisioningand mobile device power control in IoT networksrdquo IEEETransactions on Network and Service Management vol 16no 1 pp 167ndash175 2018

[35] W Wang Y Jiang and W Wu ldquoMultiagent-based resourceallocation for energy minimization in cloud computing sys-temsrdquo IEEE Transactions on Systems Man and CyberneticsSystems vol 47 no 2 pp 1ndash16 2016

[36] K Krishnajyothi ldquoParallel data processing for effective dy-namic resource allocation in the cloudrdquo International Journalof Computer Applications vol 70 no 22 pp 1ndash4 2013

[37] M M Hassan B Song M S Hossain and A Alamri ldquoEf-ficient resource scheduling for big data processing in cloudplatformrdquo in Proceedings of the International Conference onInternet and Distributed Computing Systems pp 51ndash63Springer Calabria Italy September 2014

[38] C-M Wu R-S Chang and H-Y Chan ldquoA green energy-efficient scheduling algorithm using the dvfs technique forcloud datacentersrdquo Future Generation Computer Systemsvol 37 no 7 pp 141ndash147 2014

[39] Z Wang and X Su ldquoDynamically hierarchical resource-al-location algorithm in cloud computing environmentrdquo AeJournal of Supercomputing vol 71 no 7 pp 2748ndash2766 2015

[40] W-Y Lin G-Y Lin and H-Y Wei ldquoDynamic auctionmechanism for cloud resource allocationrdquo in Proceedings ofthe 2010 10th IEEEACM International Conference on ClusterCloud and Grid Computing pp 591-592 IEEE ComputerSociety Victoria Australia May 2010

[41] Y O Yazir C Matthews R Farahbod et al ldquoDynamic re-source allocation in computing clouds using distributedmultiple criteria decision analysisrdquo in Proceedings of the 2010IEEE 3rd International Conference on Cloud Computingpp 91ndash98 IEEE Washington DC USA July 2010

[42] W Lin J Z Wang C Liang and D Qi ldquoA threshold-baseddynamic resource allocation scheme for cloud computingrdquoProcedia Engineering vol 23 no 5 pp 695ndash703 2011

[43] X Xu S Fu L Qi et al ldquoAn IoT-oriented data placementmethod with privacy preservation in cloud environmentrdquoJournal of Network and Computer Applications vol 124pp 148ndash157 2018

[44] J Bokyun P Md Jalil L Daeho and S Doug Young ldquoEf-ficient computation offloading in mobile cloud computing forvideo streaming over 5 grdquo Computers Materials amp Continuavol 61 no 2 pp 439ndash463 2019

[45] H Sarbazi-Azad and A Y Zomaya ldquoEnergy-efficient resourceutilization in cloud computingrdquo in Large Scale Network-Centric Distributed Systems pp 377ndash408 Wiley-IEEE PressHoboken NJ USA 2014

[46] C-H Hsu K D Slagter S-C Chen and Y-C ChungldquoOptimizing energy consumption with task consolidation incloudsrdquo Information Sciences vol 258 no 3 pp 452ndash4622014

[47] C-H Hsu S-C Chen C-C Lee et al ldquoEnergy-aware taskconsolidation technique for cloud computingrdquo in Proceedingsof the 2011 IEEE Aird International Conference on CloudComputing Technology and Science pp 115ndash121 IEEE Ath-ens Greece 2011

[48] S K Panda and P K Jana ldquoAn efficient task consolidationalgorithm for cloud computing systemsrdquo in Proceedings of theInternational Conference on Distributed Computing and In-ternet Technology pp 61ndash74 Springer Bhubaneswar IndiaJanuary 2016

[49] L Yin J Luo and H Luo ldquoTasks scheduling and resourceallocation in fog computing based on containers for smartmanufacturingrdquo IEEE Transactions on Industrial Informaticsvol 14 no 10 pp 4712ndash4721 2018

[50] M Aazam and E-N Huh ldquoDynamic resource provisioningthrough fog micro datacenterrdquo in Proceedings of the 2015IEEE International Conference on Pervasive Computing andCommunication Workshops (PerCom Workshops) pp 105ndash110 IEEE St Louis MO USA March 2015

[51] B Jia H Hu Y Zeng T Xu and Y Yang ldquoDouble-matchingresource allocation strategy in fog computing networks basedon cost efficiencyrdquo Journal of Communications and Networksvol 20 no 3 pp 237ndash246 2018

[52] H Zhang Y Xiao S Bu D Niyato F R Yu and Z HanldquoComputing resource allocation in three-tier IoT fog net-works a joint optimization approach combining stackelberggame and matchingrdquo IEEE Internet of Aings Journal vol 4no 5 pp 1204ndash1215 2017

[53] J Tan T-H Chang and T Q Quelc ldquoMinimum energyresource allocation in fog radio access network with fronthauland latency constraintsrdquo in Proceedings of the 2018 IEEE 19thInternational Workshop on Signal Processing Advances inWireless Communications (SPAWC) pp 1ndash5 IEEE KalamataGreece June 2018

[54] D R de Vasconcelos R M de Castro Andrade andJ N de Souza ldquoSmart shadowndashan autonomous availabilitycomputation resource allocation platform for internet ofthings in the fog computing environmentrdquo in Proceedings of the2015 International Conference on Distributed Computing inSensor Systems pp 216-217 IEEE Fortaleza Brazil June 2015

[55] M Aazam M St-Hilaire C-H Lung and I LambadarisldquoPre-fog IoT trace based probabilistic resource estimation atfogrdquo in Proceedings of the 2016 13th IEEE Annual ConsumerCommunications amp Networking Conference (CCNC) pp 12ndash17 IEEE Las Vegas NV USA January 2016

[56] N D Tung L L Bao and B Vijay ldquoPrice-based resourceallocation for edge computing a market equilibrium ap-proachrdquo IEEE Transactions on Cloud Computing p 1 2018

[57] X Xu C He Z Xu L Qi S Wan and M Z A BhuiyanldquoJoint optimization of offloading utility and privacy for edgecomputing enabled IoTrdquo IEEE Internet of Aings Journal2019

[58] X Xu Y Chen X Zhang Q Liu X Liu and L Qi ldquoAblockchain-based computation offloading method for edgecomputing in 5G networksrdquo Software Practice and Experi-ence 2019

[59] X Xu Y Li T Huang et al ldquoAn energy-aware computationoffloading method for smart edge computing in wirelessmetropolitan area networksrdquo Journal of Network and Com-puter Applications vol 133 pp 75ndash85 2019

[60] X Xu Y Xue L Qi et al ldquoAn edge computing-enabledcomputation offloading method with privacy preservation forinternet of connected vehiclesrdquo Future Generation ComputerSystems vol 96 pp 89ndash100 2019

14 Complexity

[61] G Yeting L Fang X Nong and C Zhengguo ldquoTask-basedresource allocation bid in edge computing micro datacenterrdquoComputers Materials amp Continua vol 61 no 2 pp 777ndash7922019

[62] X Chen L Jiao W Li and X Fu ldquoEfficient multi-usercomputation offloading for mobile-edge cloud computingrdquoIEEEACM Transactions on Networking vol 24 no 5pp 2795ndash2808 2016

[63] B Gao L He X Lu C Chang K Li and K Li ldquoDevelopingenergy-aware task allocation schemes in cloud-assisted mo-bile workflowsrdquo in Proceedings of the 2015 IEEE InternationalConference on Computer and Information Technology Ubiq-uitous Computing and Communications Dependable Auto-nomic and Secure Computing Pervasive Intelligence andComputing pp 1266ndash1273 IEEE Liverpool UK October2015

[64] X Xu X Zhang H Gao Y Xue L Qi andW Dou ldquoBecomeblockchain-enabled computation offloading for IoT in mobileedge computingrdquo IEEE Transactions on Industrial Infor-matics 2019

[65] W Yifei W Zhaoying G Da and Y F Richard ldquoDeepq-learning based computation offloading strategy for mobileedge computingrdquo Computers Materials amp Continua vol 59no 1 pp 89ndash104 2019

[66] M Barcelo A Correa J Llorca A M Tulino J L Vicarioand A Morell ldquoIoT-cloud service optimization in nextgeneration smart environmentsrdquo IEEE Journal on SelectedAreas in Communications vol 34 no 12 pp 4077ndash40902016

[67] B Cheng M Wang S Zhao Z Zhai D Zhu and J ChenldquoSituation-aware dynamic service coordination in an IoTenvironmentrdquo IEEEACM Transactions On Networkingvol 25 no 4 pp 2082ndash2095 2017

[68] V Angelakis I Avgouleas N Pappas and D Yuan ldquoFlexibleallocation of heterogeneous resources to services on an IoTdevicerdquo in Proceedings of the 2015 IEEE Conference onComputer Communications Workshops (INFOCOMWKSHPS) pp 99-100 IEEE Hong Kong China April 2015

[69] S Li Q Ni Y Sun G Min and S Al-Rubaye ldquoEnergy-ef-ficient resource allocation for industrial cyber-physical IoTsystems in 5g erardquo IEEE Transactions on Industrial Infor-matics vol 14 no 6 pp 2618ndash2628 2018

[70] X Liu Z Qin Y Gao and J A McCann ldquoResource allocationin wireless powered IoT networksrdquo IEEE Internet of AingsJournal vol 6 no 3 pp 4935ndash4945 2019

[71] W Ejaz and M Ibnkahla ldquoMulti-band spectrum sensing andresource allocation for IoT in cognitive 5g networksrdquo IEEEInternet of Aings Journal vol 5 no 1 pp 150ndash163 2017

[72] G Colistra V Pilloni and L Atzori ldquoTask allocation in groupof nodes in the IoT a consensus approachrdquo in Proceedings ofthe 2014 IEEE International Conference on Communications(ICC) pp 3848ndash3853 IEEE Sydney Australia June 2014

[73] J Li Q Sun and G Fan ldquoResource allocation for multiclassservice in IoTuplink communicationsrdquo inProceedings of the 20163rd International Conference on Systems and Informatics (ICSAI)pp 777ndash781 IEEE Shanghai China November 2016

[74] L I Zheng and K H Liu ldquoDynamic bandwidth resourceallocation algorithm in internet of things and its applicationrdquoComputer Engineering vol 38 no 17 pp 16ndash19 2012

[75] K Gai and M Qiu ldquoOptimal resource allocation using re-inforcement learning for IoT content-centric servicesrdquo Ap-plied Soft Computing vol 70 pp 12ndash21 2018

[76] X Xu W Dou X Zhang and J Chen ldquoEnreal an energy-aware resource allocation method for scientific workflow

executions in cloud environmentrdquo IEEE Transactions onCloud Computing vol 4 no 2 pp 166ndash179 2016

[77] K Bousselmi Z Brahmi and M M Gammoudi ldquoEnergyefficient partitioning and scheduling approach for scientificworkflows in the cloudrdquo in Proceedings of the 2016 IEEEInternational Conference on Services Computing (SCC)pp 146ndash154 IEEE San Francisco CA USA June 2016

[78] Y Sonia C Rachid K Hubert and G Bertrand ldquoMulti-objective approach for energy-aware workflow scheduling incloud computing environmentsrdquo Ae Scientific World Jour-nal vol 2013 Article ID 350934 13 pages 2013

[79] F Cao Efficient Scientific Workflow Scheduling in CloudEnvironment Southern Illinois University Carbondale ILUSA 2014

[80] Z Li J Ge H Hu W Song H Hu and B Luo ldquoCost andenergy aware scheduling algorithm for scientific workflowswith deadline constraint in cloudsrdquo IEEE Transactions onServices Computing vol 11 no 4 pp 713ndash726 2015

[81] M Khaleel and M M Zhu ldquoEnergy-aware job managementapproaches for workflow in cloudrdquo in Proceedings of the 2015IEEE International Conference on Cluster Computingpp 506-507 IEEE Chicago IL USA September 2015

[82] J Shi J Luo F Dong and J Zhang ldquoA budget and deadlineaware scientific workflow resource provisioning and sched-uling mechanism for cloudrdquo in Proceedings of the 2014 IEEE18th International Conference on Computer Supported Co-operative Work in Design (CSCWD) pp 672ndash677 IEEEHsinchu Taiwan January 2014

[83] Y Ge Y Zhang Q Qiu and Y-H Lu ldquoA game theoreticresource allocation for overall energy minimization in mobilecloud computing systemrdquo in Proceedings of the 2012ACMIEEE International Symposium on Low Power Electronicsand Design pp 279ndash284 ACM Redondo Beach CA USAAugust 2012

[84] X Wang Y Wang and C Yue ldquoAn energy-aware bi-leveloptimization model for multi-job scheduling problems undercloud computingrdquo Soft Computing vol 20 no 1 pp 303ndash3172016

[85] R Yanggratoke F Wuhib and R Stadler ldquoGossip-basedresource allocation for green computing in large cloudsrdquo inProceedings of the 2011 7th International Conference onNetwork and Service Management pp 1ndash9 IEEE ParisFrance October 2011

[86] A Paya and D C Marinescu ldquoEnergy-aware load balancingand application scaling for the cloud ecosystemrdquo IEEETransactions on Cloud Computing vol 5 no 1 pp 15ndash272015

[87] V D Justafort R Beaubrun and S Pierre ldquoA hybrid ap-proach for optimizing carbon footprint in intercloud envi-ronmentrdquo IEEE Transactions on Services Computing vol 12no 2 pp 186ndash198 2016

[88] R Panwar and B Mallick ldquoLoad balancing in cloud com-puting using dynamic load management algorithmrdquo inProceedings of the 2015 International Conference on GreenComputing and Internet of Aings (ICGCIoT) pp 773ndash778IEEE Noida India October 2015

[89] C-T Yang K-C Wang H-Y Cheng C-T Kuo andW C C Chu ldquoGreen power management with dynamicresource allocation for cloud virtual machinesrdquo in Proceedingsof the 2011 IEEE International Conference on High Perfor-mance Computing and Communications pp 726ndash733 IEEEAlberta Canada September 2011

[90] C-T Yang H-Y Cheng and K-L Huang ldquoA dynamicresource allocation model for virtual machine management

Complexity 15

on cloudrdquo in Proceedings of the International Conference onGrid and Distributed Computing pp 581ndash590 SpringerGangneug Korea December 2011

[91] X Xu F Shucun C Qing et al ldquoDynamic resource allocationfor load balancing in fog environmentrdquo Wireless Commu-nications amp Mobile Computing vol 2018 no 2 Article ID6421607 15 pages 2018

[92] J Oueis E C Strinati and S Barbarossa ldquoe fog balancingload distribution for small cell cloud computingrdquo in Pro-ceedings of the 2015 IEEE 81st Vehicular Technology Confer-ence (VTC Spring) pp 1ndash6 IEEE Glasgow UK May 2015

[93] K Wang Y Wang Y Sun S Guo and J Wu ldquoGreen in-dustrial internet of things architecture an energy-efficientperspectiverdquo IEEE Communications Magazine vol 54 no 12pp 48ndash54 2016

16 Complexity

References

[1] Q Zhang L Cheng and R Boutaba ldquoCloud computingstate-of-the-art and research challengesrdquo Journal of InternetServices and Applications vol 1 no 1 pp 7ndash18 2010

[2] W Shi J Cao Q Zhang Y Li and L Xu ldquoEdge computingvision and challengesrdquo IEEE Internet of Aings Journal vol 3no 5 pp 637ndash646 2016

[3] M Iorga L Feldman R Barton M J Martin N S Gorenand C Mahmoudi ldquoFog computing conceptual modelrdquo TechRep Recommendations of the National Institute of Standardsand Technology Gaithersburg MD USA 2018

[4] Nebbiolo ldquoFog vs edge computingrdquo Tech Rep NebbioloTechnologies Inc Milpitas CA USA 2018

[5] C T Do N H Tran C Pham M G R Alam J H Son andC S Hong ldquoA proximal algorithm for joint resource allo-cation andminimizing carbon footprint in geo-distributed fogcomputingrdquo in Proceedings of the 2015 International Con-ference on Information Networking (ICOIN) pp 324ndash329IEEE Siem Reap Cambodia January 2015

[6] S Vashi J Ram J Modi S Verma and C Prakash ldquoInternetof things (IoT) a vision architectural elements and securityissuesrdquo in Proceedings of the 2017 International Conference onI-SMAC (IoT in Social Mobile Analytics and Cloud)(I-SMAC) pp 492ndash496 IEEE Coimbatore India February2017

[7] M Mazzucco D Dyachuk and R Deters ldquoMaximizing cloudprovidersrsquo revenues via energy aware allocation policiesrdquo inProceedings of the 2010 IEEE 3rd International Conference onCloud Computing pp 131ndash138 IEEE Miami FL USA July2010

[8] Q He R Zhou X Zhang et al ldquoKeyword search for buildingservice-based systemsrdquo IEEE Transactions on Software En-gineering vol 43 no 7 pp 658ndash674 2016

[9] L Sun H Dong O K Hussain F K Hussain and A X LiuldquoA framework of cloud service selection with criteria inter-actionsrdquo Future Generation Computer Systems vol 94pp 749ndash764 2019

[10] M Mazzucco and D Dyachuk ldquoOptimizing cloud providersrevenues via energy efficient server allocationrdquo SustainableComputing Informatics and Systems vol 2 no 1 pp 1ndash122012

[11] Q He J Han F Chen et al ldquoQos-aware service selection forcustomisable multi-tenant service-based systems maturityand approachesrdquo in Proceedings of the 2015 IEEE 8th Inter-national Conference on Cloud Computing pp 237ndash244 IEEENew York NY USA July 2015

[12] L Sun J Ma H Wang Y Zhang and J Yong ldquoCloud servicedescription model an extension of usdl for cloud servicesrdquoIEEE Transactions on Services Computing vol 11 no 2pp 354ndash368 2015

[13] Y Wang Q He D Ye and Y Yang ldquoFormulating criticality-based cost-effective fault tolerance strategies for multi-tenantservice-based systemsrdquo IEEE Transactions on Software En-gineering vol 44 no 3 pp 291ndash307 2017

[14] S Mustafa K Bilal S U R Malik and S A Madani ldquoSla-aware energy efficient resource management for cloud en-vironmentsrdquo IEEE Access vol 6 pp 15004ndash15020 2018

[15] J Bi H Yuan M Tie and W Tan ldquoSla-based optimisation ofvirtualised resource for multi-tier web applications in clouddata centresrdquo Enterprise Information Systems vol 9 no 7pp 743ndash767 2015

[16] S Singh I Chana and R Buyya ldquoStar sla-aware autonomicmanagement of cloud resourcesrdquo IEEE Transactions on CloudComputing p 1 2017

[17] A Beloglazov and R Buyya ldquoEnergy efficient resourcemanagement in virtualized cloud data centersrdquo in Proceedingsof the 2010 10th IEEEACM International Conference onCluster Cloud and Grid Computing pp 826ndash831 IEEEComputer Society Melbourne Australia May 2010

[18] M Guazzone C Anglano and M Canonico ldquoEnergy-effi-cient resource management for cloud computing infra-structuresrdquo in Proceedings of the 2011 IEEE AirdInternational Conference on Cloud Computing Technology andScience pp 424ndash431 IEEE Athens Greece November 2011

[19] Y Sun J White and S Eade ldquoA model-based system toautomate cloud resource allocation and optimizationrdquo inProceedings of the International Conference on Model DrivenEngineering Languages and Systems pp 18ndash34 SpringerValencia Spain October 2014

[20] G Siddesh and K Srinivasa ldquoSla-driven dynamic resourceallocation on cloudsrdquo in Proceedings of the InternationalConference on Advanced Computing Networking and Securitypp 9ndash18 Springer Surathkal India December 2011

[21] S K Garg S K Gopalaiyengar and R Buyya ldquoSla-basedresource provisioning for heterogeneous workloads in avirtualized cloud datacenterrdquo in Proceedings of the Interna-tional Conference on Algorithms and Architectures for ParallelProcessing pp 371ndash384 Springer Melbourne AustraliaOctober 2011

[22] J Bi Z Zhu and H Yuan ldquoSla-aware dynamic resourceprovisioning for profit maximization in shared cloud datacentersrdquo in Proceedings of the International Conference onHigh Performance Networking Computing and Communi-cation Systems pp 366ndash372 Springer Singapore May 2011

[23] L Qi Y Chen Y Yuan S Fu X Zhang and X Xu ldquoA Qos-aware virtual machine scheduling method for energy con-servation in cloud-based cyber-physical systemsrdquo WorldWide Web vol 4 no 3 pp 1ndash23 2019

[24] A Beloglazov J Abawajy and R Buyya ldquoEnergy-aware re-source allocation heuristics for efficient management of datacenters for cloud computingrdquo Future Generation ComputerSystems vol 28 no 5 pp 755ndash768 2012

[25] L Qi R Wang C Hu S Li Q He and X Xu ldquoTime-awaredistributed service recommendation with privacy-preserva-tionrdquo Information Sciences vol 480 pp 354ndash364 2019

[26] Z Zhai B Cheng Y Tian J Chen L Zhao and M Niu ldquoAdata-driven service creation approach for end-usersrdquo IEEEAccess vol 4 pp 9923ndash9940 2016

[27] L Gu D Zeng S Guo A Barnawi and Y Xiang ldquoCostefficient resource management in fog computing supportedmedical cyber-physical systemrdquo IEEE Transactions onEmerging Topics in Computing vol 5 no 1 pp 108ndash119 2015

[28] L Ni J Zhang C Jiang C Yan and K Yu ldquoResource al-location strategy in fog computing based on priced timed petrinetsrdquo Ieee Internet of Aings Journal vol 4 no 5pp 1216ndash1228 2017

[29] L Wei T Yang F C Delicato et al ldquoOn enabling sustainableedge computing with renewable energy resourcesrdquo IEEECommunications Magazine vol 56 no 5 pp 94ndash101 2018

[30] P Lai Q He G Cui et al ldquoEdge user allocation with dynamicquality of servicerdquo in Proceedings of the International Con-ference on Service-Oriented Computing pp 86ndash101 SpringerToulouse France October 2019

[31] X Xu Q Liu X Zhang J Zhang L Qi and W Dou ldquoAblockchain-powered crowdsourcing method with privacy

Complexity 13

preservation in mobile environmentrdquo IEEE Transactions onComputational Social Systems vol 6 no 6 pp 1407ndash1419 2019

[32] O Rolik E Zharikov and S Telenyk ldquoMicrocloud-basedarchitecture of management system for IoT infrastructuresrdquoin Proceedings of the 2016 Aird International Scientific-Practical Conference Problems of Infocommunications Scienceand Technology (PIC SampT) pp 149ndash151 IEEE KharkivUkaraine October 2016

[33] W He S Guo Y Liang R Ma X Qiu and L Shi ldquoQos-awareand resource-efficient dynamic slicingmechanism for internetof thingsrdquo Computers Materials amp Continua vol 61 no 3pp 1345ndash1364 2019

[34] J Yao and N Ansari ldquoQos-aware fog resource provisioningand mobile device power control in IoT networksrdquo IEEETransactions on Network and Service Management vol 16no 1 pp 167ndash175 2018

[35] W Wang Y Jiang and W Wu ldquoMultiagent-based resourceallocation for energy minimization in cloud computing sys-temsrdquo IEEE Transactions on Systems Man and CyberneticsSystems vol 47 no 2 pp 1ndash16 2016

[36] K Krishnajyothi ldquoParallel data processing for effective dy-namic resource allocation in the cloudrdquo International Journalof Computer Applications vol 70 no 22 pp 1ndash4 2013

[37] M M Hassan B Song M S Hossain and A Alamri ldquoEf-ficient resource scheduling for big data processing in cloudplatformrdquo in Proceedings of the International Conference onInternet and Distributed Computing Systems pp 51ndash63Springer Calabria Italy September 2014

[38] C-M Wu R-S Chang and H-Y Chan ldquoA green energy-efficient scheduling algorithm using the dvfs technique forcloud datacentersrdquo Future Generation Computer Systemsvol 37 no 7 pp 141ndash147 2014

[39] Z Wang and X Su ldquoDynamically hierarchical resource-al-location algorithm in cloud computing environmentrdquo AeJournal of Supercomputing vol 71 no 7 pp 2748ndash2766 2015

[40] W-Y Lin G-Y Lin and H-Y Wei ldquoDynamic auctionmechanism for cloud resource allocationrdquo in Proceedings ofthe 2010 10th IEEEACM International Conference on ClusterCloud and Grid Computing pp 591-592 IEEE ComputerSociety Victoria Australia May 2010

[41] Y O Yazir C Matthews R Farahbod et al ldquoDynamic re-source allocation in computing clouds using distributedmultiple criteria decision analysisrdquo in Proceedings of the 2010IEEE 3rd International Conference on Cloud Computingpp 91ndash98 IEEE Washington DC USA July 2010

[42] W Lin J Z Wang C Liang and D Qi ldquoA threshold-baseddynamic resource allocation scheme for cloud computingrdquoProcedia Engineering vol 23 no 5 pp 695ndash703 2011

[43] X Xu S Fu L Qi et al ldquoAn IoT-oriented data placementmethod with privacy preservation in cloud environmentrdquoJournal of Network and Computer Applications vol 124pp 148ndash157 2018

[44] J Bokyun P Md Jalil L Daeho and S Doug Young ldquoEf-ficient computation offloading in mobile cloud computing forvideo streaming over 5 grdquo Computers Materials amp Continuavol 61 no 2 pp 439ndash463 2019

[45] H Sarbazi-Azad and A Y Zomaya ldquoEnergy-efficient resourceutilization in cloud computingrdquo in Large Scale Network-Centric Distributed Systems pp 377ndash408 Wiley-IEEE PressHoboken NJ USA 2014

[46] C-H Hsu K D Slagter S-C Chen and Y-C ChungldquoOptimizing energy consumption with task consolidation incloudsrdquo Information Sciences vol 258 no 3 pp 452ndash4622014

[47] C-H Hsu S-C Chen C-C Lee et al ldquoEnergy-aware taskconsolidation technique for cloud computingrdquo in Proceedingsof the 2011 IEEE Aird International Conference on CloudComputing Technology and Science pp 115ndash121 IEEE Ath-ens Greece 2011

[48] S K Panda and P K Jana ldquoAn efficient task consolidationalgorithm for cloud computing systemsrdquo in Proceedings of theInternational Conference on Distributed Computing and In-ternet Technology pp 61ndash74 Springer Bhubaneswar IndiaJanuary 2016

[49] L Yin J Luo and H Luo ldquoTasks scheduling and resourceallocation in fog computing based on containers for smartmanufacturingrdquo IEEE Transactions on Industrial Informaticsvol 14 no 10 pp 4712ndash4721 2018

[50] M Aazam and E-N Huh ldquoDynamic resource provisioningthrough fog micro datacenterrdquo in Proceedings of the 2015IEEE International Conference on Pervasive Computing andCommunication Workshops (PerCom Workshops) pp 105ndash110 IEEE St Louis MO USA March 2015

[51] B Jia H Hu Y Zeng T Xu and Y Yang ldquoDouble-matchingresource allocation strategy in fog computing networks basedon cost efficiencyrdquo Journal of Communications and Networksvol 20 no 3 pp 237ndash246 2018

[52] H Zhang Y Xiao S Bu D Niyato F R Yu and Z HanldquoComputing resource allocation in three-tier IoT fog net-works a joint optimization approach combining stackelberggame and matchingrdquo IEEE Internet of Aings Journal vol 4no 5 pp 1204ndash1215 2017

[53] J Tan T-H Chang and T Q Quelc ldquoMinimum energyresource allocation in fog radio access network with fronthauland latency constraintsrdquo in Proceedings of the 2018 IEEE 19thInternational Workshop on Signal Processing Advances inWireless Communications (SPAWC) pp 1ndash5 IEEE KalamataGreece June 2018

[54] D R de Vasconcelos R M de Castro Andrade andJ N de Souza ldquoSmart shadowndashan autonomous availabilitycomputation resource allocation platform for internet ofthings in the fog computing environmentrdquo in Proceedings of the2015 International Conference on Distributed Computing inSensor Systems pp 216-217 IEEE Fortaleza Brazil June 2015

[55] M Aazam M St-Hilaire C-H Lung and I LambadarisldquoPre-fog IoT trace based probabilistic resource estimation atfogrdquo in Proceedings of the 2016 13th IEEE Annual ConsumerCommunications amp Networking Conference (CCNC) pp 12ndash17 IEEE Las Vegas NV USA January 2016

[56] N D Tung L L Bao and B Vijay ldquoPrice-based resourceallocation for edge computing a market equilibrium ap-proachrdquo IEEE Transactions on Cloud Computing p 1 2018

[57] X Xu C He Z Xu L Qi S Wan and M Z A BhuiyanldquoJoint optimization of offloading utility and privacy for edgecomputing enabled IoTrdquo IEEE Internet of Aings Journal2019

[58] X Xu Y Chen X Zhang Q Liu X Liu and L Qi ldquoAblockchain-based computation offloading method for edgecomputing in 5G networksrdquo Software Practice and Experi-ence 2019

[59] X Xu Y Li T Huang et al ldquoAn energy-aware computationoffloading method for smart edge computing in wirelessmetropolitan area networksrdquo Journal of Network and Com-puter Applications vol 133 pp 75ndash85 2019

[60] X Xu Y Xue L Qi et al ldquoAn edge computing-enabledcomputation offloading method with privacy preservation forinternet of connected vehiclesrdquo Future Generation ComputerSystems vol 96 pp 89ndash100 2019

14 Complexity

[61] G Yeting L Fang X Nong and C Zhengguo ldquoTask-basedresource allocation bid in edge computing micro datacenterrdquoComputers Materials amp Continua vol 61 no 2 pp 777ndash7922019

[62] X Chen L Jiao W Li and X Fu ldquoEfficient multi-usercomputation offloading for mobile-edge cloud computingrdquoIEEEACM Transactions on Networking vol 24 no 5pp 2795ndash2808 2016

[63] B Gao L He X Lu C Chang K Li and K Li ldquoDevelopingenergy-aware task allocation schemes in cloud-assisted mo-bile workflowsrdquo in Proceedings of the 2015 IEEE InternationalConference on Computer and Information Technology Ubiq-uitous Computing and Communications Dependable Auto-nomic and Secure Computing Pervasive Intelligence andComputing pp 1266ndash1273 IEEE Liverpool UK October2015

[64] X Xu X Zhang H Gao Y Xue L Qi andW Dou ldquoBecomeblockchain-enabled computation offloading for IoT in mobileedge computingrdquo IEEE Transactions on Industrial Infor-matics 2019

[65] W Yifei W Zhaoying G Da and Y F Richard ldquoDeepq-learning based computation offloading strategy for mobileedge computingrdquo Computers Materials amp Continua vol 59no 1 pp 89ndash104 2019

[66] M Barcelo A Correa J Llorca A M Tulino J L Vicarioand A Morell ldquoIoT-cloud service optimization in nextgeneration smart environmentsrdquo IEEE Journal on SelectedAreas in Communications vol 34 no 12 pp 4077ndash40902016

[67] B Cheng M Wang S Zhao Z Zhai D Zhu and J ChenldquoSituation-aware dynamic service coordination in an IoTenvironmentrdquo IEEEACM Transactions On Networkingvol 25 no 4 pp 2082ndash2095 2017

[68] V Angelakis I Avgouleas N Pappas and D Yuan ldquoFlexibleallocation of heterogeneous resources to services on an IoTdevicerdquo in Proceedings of the 2015 IEEE Conference onComputer Communications Workshops (INFOCOMWKSHPS) pp 99-100 IEEE Hong Kong China April 2015

[69] S Li Q Ni Y Sun G Min and S Al-Rubaye ldquoEnergy-ef-ficient resource allocation for industrial cyber-physical IoTsystems in 5g erardquo IEEE Transactions on Industrial Infor-matics vol 14 no 6 pp 2618ndash2628 2018

[70] X Liu Z Qin Y Gao and J A McCann ldquoResource allocationin wireless powered IoT networksrdquo IEEE Internet of AingsJournal vol 6 no 3 pp 4935ndash4945 2019

[71] W Ejaz and M Ibnkahla ldquoMulti-band spectrum sensing andresource allocation for IoT in cognitive 5g networksrdquo IEEEInternet of Aings Journal vol 5 no 1 pp 150ndash163 2017

[72] G Colistra V Pilloni and L Atzori ldquoTask allocation in groupof nodes in the IoT a consensus approachrdquo in Proceedings ofthe 2014 IEEE International Conference on Communications(ICC) pp 3848ndash3853 IEEE Sydney Australia June 2014

[73] J Li Q Sun and G Fan ldquoResource allocation for multiclassservice in IoTuplink communicationsrdquo inProceedings of the 20163rd International Conference on Systems and Informatics (ICSAI)pp 777ndash781 IEEE Shanghai China November 2016

[74] L I Zheng and K H Liu ldquoDynamic bandwidth resourceallocation algorithm in internet of things and its applicationrdquoComputer Engineering vol 38 no 17 pp 16ndash19 2012

[75] K Gai and M Qiu ldquoOptimal resource allocation using re-inforcement learning for IoT content-centric servicesrdquo Ap-plied Soft Computing vol 70 pp 12ndash21 2018

[76] X Xu W Dou X Zhang and J Chen ldquoEnreal an energy-aware resource allocation method for scientific workflow

executions in cloud environmentrdquo IEEE Transactions onCloud Computing vol 4 no 2 pp 166ndash179 2016

[77] K Bousselmi Z Brahmi and M M Gammoudi ldquoEnergyefficient partitioning and scheduling approach for scientificworkflows in the cloudrdquo in Proceedings of the 2016 IEEEInternational Conference on Services Computing (SCC)pp 146ndash154 IEEE San Francisco CA USA June 2016

[78] Y Sonia C Rachid K Hubert and G Bertrand ldquoMulti-objective approach for energy-aware workflow scheduling incloud computing environmentsrdquo Ae Scientific World Jour-nal vol 2013 Article ID 350934 13 pages 2013

[79] F Cao Efficient Scientific Workflow Scheduling in CloudEnvironment Southern Illinois University Carbondale ILUSA 2014

[80] Z Li J Ge H Hu W Song H Hu and B Luo ldquoCost andenergy aware scheduling algorithm for scientific workflowswith deadline constraint in cloudsrdquo IEEE Transactions onServices Computing vol 11 no 4 pp 713ndash726 2015

[81] M Khaleel and M M Zhu ldquoEnergy-aware job managementapproaches for workflow in cloudrdquo in Proceedings of the 2015IEEE International Conference on Cluster Computingpp 506-507 IEEE Chicago IL USA September 2015

[82] J Shi J Luo F Dong and J Zhang ldquoA budget and deadlineaware scientific workflow resource provisioning and sched-uling mechanism for cloudrdquo in Proceedings of the 2014 IEEE18th International Conference on Computer Supported Co-operative Work in Design (CSCWD) pp 672ndash677 IEEEHsinchu Taiwan January 2014

[83] Y Ge Y Zhang Q Qiu and Y-H Lu ldquoA game theoreticresource allocation for overall energy minimization in mobilecloud computing systemrdquo in Proceedings of the 2012ACMIEEE International Symposium on Low Power Electronicsand Design pp 279ndash284 ACM Redondo Beach CA USAAugust 2012

[84] X Wang Y Wang and C Yue ldquoAn energy-aware bi-leveloptimization model for multi-job scheduling problems undercloud computingrdquo Soft Computing vol 20 no 1 pp 303ndash3172016

[85] R Yanggratoke F Wuhib and R Stadler ldquoGossip-basedresource allocation for green computing in large cloudsrdquo inProceedings of the 2011 7th International Conference onNetwork and Service Management pp 1ndash9 IEEE ParisFrance October 2011

[86] A Paya and D C Marinescu ldquoEnergy-aware load balancingand application scaling for the cloud ecosystemrdquo IEEETransactions on Cloud Computing vol 5 no 1 pp 15ndash272015

[87] V D Justafort R Beaubrun and S Pierre ldquoA hybrid ap-proach for optimizing carbon footprint in intercloud envi-ronmentrdquo IEEE Transactions on Services Computing vol 12no 2 pp 186ndash198 2016

[88] R Panwar and B Mallick ldquoLoad balancing in cloud com-puting using dynamic load management algorithmrdquo inProceedings of the 2015 International Conference on GreenComputing and Internet of Aings (ICGCIoT) pp 773ndash778IEEE Noida India October 2015

[89] C-T Yang K-C Wang H-Y Cheng C-T Kuo andW C C Chu ldquoGreen power management with dynamicresource allocation for cloud virtual machinesrdquo in Proceedingsof the 2011 IEEE International Conference on High Perfor-mance Computing and Communications pp 726ndash733 IEEEAlberta Canada September 2011

[90] C-T Yang H-Y Cheng and K-L Huang ldquoA dynamicresource allocation model for virtual machine management

Complexity 15

on cloudrdquo in Proceedings of the International Conference onGrid and Distributed Computing pp 581ndash590 SpringerGangneug Korea December 2011

[91] X Xu F Shucun C Qing et al ldquoDynamic resource allocationfor load balancing in fog environmentrdquo Wireless Commu-nications amp Mobile Computing vol 2018 no 2 Article ID6421607 15 pages 2018

[92] J Oueis E C Strinati and S Barbarossa ldquoe fog balancingload distribution for small cell cloud computingrdquo in Pro-ceedings of the 2015 IEEE 81st Vehicular Technology Confer-ence (VTC Spring) pp 1ndash6 IEEE Glasgow UK May 2015

[93] K Wang Y Wang Y Sun S Guo and J Wu ldquoGreen in-dustrial internet of things architecture an energy-efficientperspectiverdquo IEEE Communications Magazine vol 54 no 12pp 48ndash54 2016

16 Complexity

preservation in mobile environmentrdquo IEEE Transactions onComputational Social Systems vol 6 no 6 pp 1407ndash1419 2019

[32] O Rolik E Zharikov and S Telenyk ldquoMicrocloud-basedarchitecture of management system for IoT infrastructuresrdquoin Proceedings of the 2016 Aird International Scientific-Practical Conference Problems of Infocommunications Scienceand Technology (PIC SampT) pp 149ndash151 IEEE KharkivUkaraine October 2016

[33] W He S Guo Y Liang R Ma X Qiu and L Shi ldquoQos-awareand resource-efficient dynamic slicingmechanism for internetof thingsrdquo Computers Materials amp Continua vol 61 no 3pp 1345ndash1364 2019

[34] J Yao and N Ansari ldquoQos-aware fog resource provisioningand mobile device power control in IoT networksrdquo IEEETransactions on Network and Service Management vol 16no 1 pp 167ndash175 2018

[35] W Wang Y Jiang and W Wu ldquoMultiagent-based resourceallocation for energy minimization in cloud computing sys-temsrdquo IEEE Transactions on Systems Man and CyberneticsSystems vol 47 no 2 pp 1ndash16 2016

[36] K Krishnajyothi ldquoParallel data processing for effective dy-namic resource allocation in the cloudrdquo International Journalof Computer Applications vol 70 no 22 pp 1ndash4 2013

[37] M M Hassan B Song M S Hossain and A Alamri ldquoEf-ficient resource scheduling for big data processing in cloudplatformrdquo in Proceedings of the International Conference onInternet and Distributed Computing Systems pp 51ndash63Springer Calabria Italy September 2014

[38] C-M Wu R-S Chang and H-Y Chan ldquoA green energy-efficient scheduling algorithm using the dvfs technique forcloud datacentersrdquo Future Generation Computer Systemsvol 37 no 7 pp 141ndash147 2014

[39] Z Wang and X Su ldquoDynamically hierarchical resource-al-location algorithm in cloud computing environmentrdquo AeJournal of Supercomputing vol 71 no 7 pp 2748ndash2766 2015

[40] W-Y Lin G-Y Lin and H-Y Wei ldquoDynamic auctionmechanism for cloud resource allocationrdquo in Proceedings ofthe 2010 10th IEEEACM International Conference on ClusterCloud and Grid Computing pp 591-592 IEEE ComputerSociety Victoria Australia May 2010

[41] Y O Yazir C Matthews R Farahbod et al ldquoDynamic re-source allocation in computing clouds using distributedmultiple criteria decision analysisrdquo in Proceedings of the 2010IEEE 3rd International Conference on Cloud Computingpp 91ndash98 IEEE Washington DC USA July 2010

[42] W Lin J Z Wang C Liang and D Qi ldquoA threshold-baseddynamic resource allocation scheme for cloud computingrdquoProcedia Engineering vol 23 no 5 pp 695ndash703 2011

[43] X Xu S Fu L Qi et al ldquoAn IoT-oriented data placementmethod with privacy preservation in cloud environmentrdquoJournal of Network and Computer Applications vol 124pp 148ndash157 2018

[44] J Bokyun P Md Jalil L Daeho and S Doug Young ldquoEf-ficient computation offloading in mobile cloud computing forvideo streaming over 5 grdquo Computers Materials amp Continuavol 61 no 2 pp 439ndash463 2019

[45] H Sarbazi-Azad and A Y Zomaya ldquoEnergy-efficient resourceutilization in cloud computingrdquo in Large Scale Network-Centric Distributed Systems pp 377ndash408 Wiley-IEEE PressHoboken NJ USA 2014

[46] C-H Hsu K D Slagter S-C Chen and Y-C ChungldquoOptimizing energy consumption with task consolidation incloudsrdquo Information Sciences vol 258 no 3 pp 452ndash4622014

[47] C-H Hsu S-C Chen C-C Lee et al ldquoEnergy-aware taskconsolidation technique for cloud computingrdquo in Proceedingsof the 2011 IEEE Aird International Conference on CloudComputing Technology and Science pp 115ndash121 IEEE Ath-ens Greece 2011

[48] S K Panda and P K Jana ldquoAn efficient task consolidationalgorithm for cloud computing systemsrdquo in Proceedings of theInternational Conference on Distributed Computing and In-ternet Technology pp 61ndash74 Springer Bhubaneswar IndiaJanuary 2016

[49] L Yin J Luo and H Luo ldquoTasks scheduling and resourceallocation in fog computing based on containers for smartmanufacturingrdquo IEEE Transactions on Industrial Informaticsvol 14 no 10 pp 4712ndash4721 2018

[50] M Aazam and E-N Huh ldquoDynamic resource provisioningthrough fog micro datacenterrdquo in Proceedings of the 2015IEEE International Conference on Pervasive Computing andCommunication Workshops (PerCom Workshops) pp 105ndash110 IEEE St Louis MO USA March 2015

[51] B Jia H Hu Y Zeng T Xu and Y Yang ldquoDouble-matchingresource allocation strategy in fog computing networks basedon cost efficiencyrdquo Journal of Communications and Networksvol 20 no 3 pp 237ndash246 2018

[52] H Zhang Y Xiao S Bu D Niyato F R Yu and Z HanldquoComputing resource allocation in three-tier IoT fog net-works a joint optimization approach combining stackelberggame and matchingrdquo IEEE Internet of Aings Journal vol 4no 5 pp 1204ndash1215 2017

[53] J Tan T-H Chang and T Q Quelc ldquoMinimum energyresource allocation in fog radio access network with fronthauland latency constraintsrdquo in Proceedings of the 2018 IEEE 19thInternational Workshop on Signal Processing Advances inWireless Communications (SPAWC) pp 1ndash5 IEEE KalamataGreece June 2018

[54] D R de Vasconcelos R M de Castro Andrade andJ N de Souza ldquoSmart shadowndashan autonomous availabilitycomputation resource allocation platform for internet ofthings in the fog computing environmentrdquo in Proceedings of the2015 International Conference on Distributed Computing inSensor Systems pp 216-217 IEEE Fortaleza Brazil June 2015

[55] M Aazam M St-Hilaire C-H Lung and I LambadarisldquoPre-fog IoT trace based probabilistic resource estimation atfogrdquo in Proceedings of the 2016 13th IEEE Annual ConsumerCommunications amp Networking Conference (CCNC) pp 12ndash17 IEEE Las Vegas NV USA January 2016

[56] N D Tung L L Bao and B Vijay ldquoPrice-based resourceallocation for edge computing a market equilibrium ap-proachrdquo IEEE Transactions on Cloud Computing p 1 2018

[57] X Xu C He Z Xu L Qi S Wan and M Z A BhuiyanldquoJoint optimization of offloading utility and privacy for edgecomputing enabled IoTrdquo IEEE Internet of Aings Journal2019

[58] X Xu Y Chen X Zhang Q Liu X Liu and L Qi ldquoAblockchain-based computation offloading method for edgecomputing in 5G networksrdquo Software Practice and Experi-ence 2019

[59] X Xu Y Li T Huang et al ldquoAn energy-aware computationoffloading method for smart edge computing in wirelessmetropolitan area networksrdquo Journal of Network and Com-puter Applications vol 133 pp 75ndash85 2019

[60] X Xu Y Xue L Qi et al ldquoAn edge computing-enabledcomputation offloading method with privacy preservation forinternet of connected vehiclesrdquo Future Generation ComputerSystems vol 96 pp 89ndash100 2019

14 Complexity

[61] G Yeting L Fang X Nong and C Zhengguo ldquoTask-basedresource allocation bid in edge computing micro datacenterrdquoComputers Materials amp Continua vol 61 no 2 pp 777ndash7922019

[62] X Chen L Jiao W Li and X Fu ldquoEfficient multi-usercomputation offloading for mobile-edge cloud computingrdquoIEEEACM Transactions on Networking vol 24 no 5pp 2795ndash2808 2016

[63] B Gao L He X Lu C Chang K Li and K Li ldquoDevelopingenergy-aware task allocation schemes in cloud-assisted mo-bile workflowsrdquo in Proceedings of the 2015 IEEE InternationalConference on Computer and Information Technology Ubiq-uitous Computing and Communications Dependable Auto-nomic and Secure Computing Pervasive Intelligence andComputing pp 1266ndash1273 IEEE Liverpool UK October2015

[64] X Xu X Zhang H Gao Y Xue L Qi andW Dou ldquoBecomeblockchain-enabled computation offloading for IoT in mobileedge computingrdquo IEEE Transactions on Industrial Infor-matics 2019

[65] W Yifei W Zhaoying G Da and Y F Richard ldquoDeepq-learning based computation offloading strategy for mobileedge computingrdquo Computers Materials amp Continua vol 59no 1 pp 89ndash104 2019

[66] M Barcelo A Correa J Llorca A M Tulino J L Vicarioand A Morell ldquoIoT-cloud service optimization in nextgeneration smart environmentsrdquo IEEE Journal on SelectedAreas in Communications vol 34 no 12 pp 4077ndash40902016

[67] B Cheng M Wang S Zhao Z Zhai D Zhu and J ChenldquoSituation-aware dynamic service coordination in an IoTenvironmentrdquo IEEEACM Transactions On Networkingvol 25 no 4 pp 2082ndash2095 2017

[68] V Angelakis I Avgouleas N Pappas and D Yuan ldquoFlexibleallocation of heterogeneous resources to services on an IoTdevicerdquo in Proceedings of the 2015 IEEE Conference onComputer Communications Workshops (INFOCOMWKSHPS) pp 99-100 IEEE Hong Kong China April 2015

[69] S Li Q Ni Y Sun G Min and S Al-Rubaye ldquoEnergy-ef-ficient resource allocation for industrial cyber-physical IoTsystems in 5g erardquo IEEE Transactions on Industrial Infor-matics vol 14 no 6 pp 2618ndash2628 2018

[70] X Liu Z Qin Y Gao and J A McCann ldquoResource allocationin wireless powered IoT networksrdquo IEEE Internet of AingsJournal vol 6 no 3 pp 4935ndash4945 2019

[71] W Ejaz and M Ibnkahla ldquoMulti-band spectrum sensing andresource allocation for IoT in cognitive 5g networksrdquo IEEEInternet of Aings Journal vol 5 no 1 pp 150ndash163 2017

[72] G Colistra V Pilloni and L Atzori ldquoTask allocation in groupof nodes in the IoT a consensus approachrdquo in Proceedings ofthe 2014 IEEE International Conference on Communications(ICC) pp 3848ndash3853 IEEE Sydney Australia June 2014

[73] J Li Q Sun and G Fan ldquoResource allocation for multiclassservice in IoTuplink communicationsrdquo inProceedings of the 20163rd International Conference on Systems and Informatics (ICSAI)pp 777ndash781 IEEE Shanghai China November 2016

[74] L I Zheng and K H Liu ldquoDynamic bandwidth resourceallocation algorithm in internet of things and its applicationrdquoComputer Engineering vol 38 no 17 pp 16ndash19 2012

[75] K Gai and M Qiu ldquoOptimal resource allocation using re-inforcement learning for IoT content-centric servicesrdquo Ap-plied Soft Computing vol 70 pp 12ndash21 2018

[76] X Xu W Dou X Zhang and J Chen ldquoEnreal an energy-aware resource allocation method for scientific workflow

executions in cloud environmentrdquo IEEE Transactions onCloud Computing vol 4 no 2 pp 166ndash179 2016

[77] K Bousselmi Z Brahmi and M M Gammoudi ldquoEnergyefficient partitioning and scheduling approach for scientificworkflows in the cloudrdquo in Proceedings of the 2016 IEEEInternational Conference on Services Computing (SCC)pp 146ndash154 IEEE San Francisco CA USA June 2016

[78] Y Sonia C Rachid K Hubert and G Bertrand ldquoMulti-objective approach for energy-aware workflow scheduling incloud computing environmentsrdquo Ae Scientific World Jour-nal vol 2013 Article ID 350934 13 pages 2013

[79] F Cao Efficient Scientific Workflow Scheduling in CloudEnvironment Southern Illinois University Carbondale ILUSA 2014

[80] Z Li J Ge H Hu W Song H Hu and B Luo ldquoCost andenergy aware scheduling algorithm for scientific workflowswith deadline constraint in cloudsrdquo IEEE Transactions onServices Computing vol 11 no 4 pp 713ndash726 2015

[81] M Khaleel and M M Zhu ldquoEnergy-aware job managementapproaches for workflow in cloudrdquo in Proceedings of the 2015IEEE International Conference on Cluster Computingpp 506-507 IEEE Chicago IL USA September 2015

[82] J Shi J Luo F Dong and J Zhang ldquoA budget and deadlineaware scientific workflow resource provisioning and sched-uling mechanism for cloudrdquo in Proceedings of the 2014 IEEE18th International Conference on Computer Supported Co-operative Work in Design (CSCWD) pp 672ndash677 IEEEHsinchu Taiwan January 2014

[83] Y Ge Y Zhang Q Qiu and Y-H Lu ldquoA game theoreticresource allocation for overall energy minimization in mobilecloud computing systemrdquo in Proceedings of the 2012ACMIEEE International Symposium on Low Power Electronicsand Design pp 279ndash284 ACM Redondo Beach CA USAAugust 2012

[84] X Wang Y Wang and C Yue ldquoAn energy-aware bi-leveloptimization model for multi-job scheduling problems undercloud computingrdquo Soft Computing vol 20 no 1 pp 303ndash3172016

[85] R Yanggratoke F Wuhib and R Stadler ldquoGossip-basedresource allocation for green computing in large cloudsrdquo inProceedings of the 2011 7th International Conference onNetwork and Service Management pp 1ndash9 IEEE ParisFrance October 2011

[86] A Paya and D C Marinescu ldquoEnergy-aware load balancingand application scaling for the cloud ecosystemrdquo IEEETransactions on Cloud Computing vol 5 no 1 pp 15ndash272015

[87] V D Justafort R Beaubrun and S Pierre ldquoA hybrid ap-proach for optimizing carbon footprint in intercloud envi-ronmentrdquo IEEE Transactions on Services Computing vol 12no 2 pp 186ndash198 2016

[88] R Panwar and B Mallick ldquoLoad balancing in cloud com-puting using dynamic load management algorithmrdquo inProceedings of the 2015 International Conference on GreenComputing and Internet of Aings (ICGCIoT) pp 773ndash778IEEE Noida India October 2015

[89] C-T Yang K-C Wang H-Y Cheng C-T Kuo andW C C Chu ldquoGreen power management with dynamicresource allocation for cloud virtual machinesrdquo in Proceedingsof the 2011 IEEE International Conference on High Perfor-mance Computing and Communications pp 726ndash733 IEEEAlberta Canada September 2011

[90] C-T Yang H-Y Cheng and K-L Huang ldquoA dynamicresource allocation model for virtual machine management

Complexity 15

on cloudrdquo in Proceedings of the International Conference onGrid and Distributed Computing pp 581ndash590 SpringerGangneug Korea December 2011

[91] X Xu F Shucun C Qing et al ldquoDynamic resource allocationfor load balancing in fog environmentrdquo Wireless Commu-nications amp Mobile Computing vol 2018 no 2 Article ID6421607 15 pages 2018

[92] J Oueis E C Strinati and S Barbarossa ldquoe fog balancingload distribution for small cell cloud computingrdquo in Pro-ceedings of the 2015 IEEE 81st Vehicular Technology Confer-ence (VTC Spring) pp 1ndash6 IEEE Glasgow UK May 2015

[93] K Wang Y Wang Y Sun S Guo and J Wu ldquoGreen in-dustrial internet of things architecture an energy-efficientperspectiverdquo IEEE Communications Magazine vol 54 no 12pp 48ndash54 2016

16 Complexity

[61] G Yeting L Fang X Nong and C Zhengguo ldquoTask-basedresource allocation bid in edge computing micro datacenterrdquoComputers Materials amp Continua vol 61 no 2 pp 777ndash7922019

[62] X Chen L Jiao W Li and X Fu ldquoEfficient multi-usercomputation offloading for mobile-edge cloud computingrdquoIEEEACM Transactions on Networking vol 24 no 5pp 2795ndash2808 2016

[63] B Gao L He X Lu C Chang K Li and K Li ldquoDevelopingenergy-aware task allocation schemes in cloud-assisted mo-bile workflowsrdquo in Proceedings of the 2015 IEEE InternationalConference on Computer and Information Technology Ubiq-uitous Computing and Communications Dependable Auto-nomic and Secure Computing Pervasive Intelligence andComputing pp 1266ndash1273 IEEE Liverpool UK October2015

[64] X Xu X Zhang H Gao Y Xue L Qi andW Dou ldquoBecomeblockchain-enabled computation offloading for IoT in mobileedge computingrdquo IEEE Transactions on Industrial Infor-matics 2019

[65] W Yifei W Zhaoying G Da and Y F Richard ldquoDeepq-learning based computation offloading strategy for mobileedge computingrdquo Computers Materials amp Continua vol 59no 1 pp 89ndash104 2019

[66] M Barcelo A Correa J Llorca A M Tulino J L Vicarioand A Morell ldquoIoT-cloud service optimization in nextgeneration smart environmentsrdquo IEEE Journal on SelectedAreas in Communications vol 34 no 12 pp 4077ndash40902016

[67] B Cheng M Wang S Zhao Z Zhai D Zhu and J ChenldquoSituation-aware dynamic service coordination in an IoTenvironmentrdquo IEEEACM Transactions On Networkingvol 25 no 4 pp 2082ndash2095 2017

[68] V Angelakis I Avgouleas N Pappas and D Yuan ldquoFlexibleallocation of heterogeneous resources to services on an IoTdevicerdquo in Proceedings of the 2015 IEEE Conference onComputer Communications Workshops (INFOCOMWKSHPS) pp 99-100 IEEE Hong Kong China April 2015

[69] S Li Q Ni Y Sun G Min and S Al-Rubaye ldquoEnergy-ef-ficient resource allocation for industrial cyber-physical IoTsystems in 5g erardquo IEEE Transactions on Industrial Infor-matics vol 14 no 6 pp 2618ndash2628 2018

[70] X Liu Z Qin Y Gao and J A McCann ldquoResource allocationin wireless powered IoT networksrdquo IEEE Internet of AingsJournal vol 6 no 3 pp 4935ndash4945 2019

[71] W Ejaz and M Ibnkahla ldquoMulti-band spectrum sensing andresource allocation for IoT in cognitive 5g networksrdquo IEEEInternet of Aings Journal vol 5 no 1 pp 150ndash163 2017

[72] G Colistra V Pilloni and L Atzori ldquoTask allocation in groupof nodes in the IoT a consensus approachrdquo in Proceedings ofthe 2014 IEEE International Conference on Communications(ICC) pp 3848ndash3853 IEEE Sydney Australia June 2014

[73] J Li Q Sun and G Fan ldquoResource allocation for multiclassservice in IoTuplink communicationsrdquo inProceedings of the 20163rd International Conference on Systems and Informatics (ICSAI)pp 777ndash781 IEEE Shanghai China November 2016

[74] L I Zheng and K H Liu ldquoDynamic bandwidth resourceallocation algorithm in internet of things and its applicationrdquoComputer Engineering vol 38 no 17 pp 16ndash19 2012

[75] K Gai and M Qiu ldquoOptimal resource allocation using re-inforcement learning for IoT content-centric servicesrdquo Ap-plied Soft Computing vol 70 pp 12ndash21 2018

[76] X Xu W Dou X Zhang and J Chen ldquoEnreal an energy-aware resource allocation method for scientific workflow

executions in cloud environmentrdquo IEEE Transactions onCloud Computing vol 4 no 2 pp 166ndash179 2016

[77] K Bousselmi Z Brahmi and M M Gammoudi ldquoEnergyefficient partitioning and scheduling approach for scientificworkflows in the cloudrdquo in Proceedings of the 2016 IEEEInternational Conference on Services Computing (SCC)pp 146ndash154 IEEE San Francisco CA USA June 2016

[78] Y Sonia C Rachid K Hubert and G Bertrand ldquoMulti-objective approach for energy-aware workflow scheduling incloud computing environmentsrdquo Ae Scientific World Jour-nal vol 2013 Article ID 350934 13 pages 2013

[79] F Cao Efficient Scientific Workflow Scheduling in CloudEnvironment Southern Illinois University Carbondale ILUSA 2014

[80] Z Li J Ge H Hu W Song H Hu and B Luo ldquoCost andenergy aware scheduling algorithm for scientific workflowswith deadline constraint in cloudsrdquo IEEE Transactions onServices Computing vol 11 no 4 pp 713ndash726 2015

[81] M Khaleel and M M Zhu ldquoEnergy-aware job managementapproaches for workflow in cloudrdquo in Proceedings of the 2015IEEE International Conference on Cluster Computingpp 506-507 IEEE Chicago IL USA September 2015

[82] J Shi J Luo F Dong and J Zhang ldquoA budget and deadlineaware scientific workflow resource provisioning and sched-uling mechanism for cloudrdquo in Proceedings of the 2014 IEEE18th International Conference on Computer Supported Co-operative Work in Design (CSCWD) pp 672ndash677 IEEEHsinchu Taiwan January 2014

[83] Y Ge Y Zhang Q Qiu and Y-H Lu ldquoA game theoreticresource allocation for overall energy minimization in mobilecloud computing systemrdquo in Proceedings of the 2012ACMIEEE International Symposium on Low Power Electronicsand Design pp 279ndash284 ACM Redondo Beach CA USAAugust 2012

[84] X Wang Y Wang and C Yue ldquoAn energy-aware bi-leveloptimization model for multi-job scheduling problems undercloud computingrdquo Soft Computing vol 20 no 1 pp 303ndash3172016

[85] R Yanggratoke F Wuhib and R Stadler ldquoGossip-basedresource allocation for green computing in large cloudsrdquo inProceedings of the 2011 7th International Conference onNetwork and Service Management pp 1ndash9 IEEE ParisFrance October 2011

[86] A Paya and D C Marinescu ldquoEnergy-aware load balancingand application scaling for the cloud ecosystemrdquo IEEETransactions on Cloud Computing vol 5 no 1 pp 15ndash272015

[87] V D Justafort R Beaubrun and S Pierre ldquoA hybrid ap-proach for optimizing carbon footprint in intercloud envi-ronmentrdquo IEEE Transactions on Services Computing vol 12no 2 pp 186ndash198 2016

[88] R Panwar and B Mallick ldquoLoad balancing in cloud com-puting using dynamic load management algorithmrdquo inProceedings of the 2015 International Conference on GreenComputing and Internet of Aings (ICGCIoT) pp 773ndash778IEEE Noida India October 2015

[89] C-T Yang K-C Wang H-Y Cheng C-T Kuo andW C C Chu ldquoGreen power management with dynamicresource allocation for cloud virtual machinesrdquo in Proceedingsof the 2011 IEEE International Conference on High Perfor-mance Computing and Communications pp 726ndash733 IEEEAlberta Canada September 2011

[90] C-T Yang H-Y Cheng and K-L Huang ldquoA dynamicresource allocation model for virtual machine management

Complexity 15

on cloudrdquo in Proceedings of the International Conference onGrid and Distributed Computing pp 581ndash590 SpringerGangneug Korea December 2011

[91] X Xu F Shucun C Qing et al ldquoDynamic resource allocationfor load balancing in fog environmentrdquo Wireless Commu-nications amp Mobile Computing vol 2018 no 2 Article ID6421607 15 pages 2018

[92] J Oueis E C Strinati and S Barbarossa ldquoe fog balancingload distribution for small cell cloud computingrdquo in Pro-ceedings of the 2015 IEEE 81st Vehicular Technology Confer-ence (VTC Spring) pp 1ndash6 IEEE Glasgow UK May 2015

[93] K Wang Y Wang Y Sun S Guo and J Wu ldquoGreen in-dustrial internet of things architecture an energy-efficientperspectiverdquo IEEE Communications Magazine vol 54 no 12pp 48ndash54 2016

16 Complexity

on cloudrdquo in Proceedings of the International Conference onGrid and Distributed Computing pp 581ndash590 SpringerGangneug Korea December 2011

[91] X Xu F Shucun C Qing et al ldquoDynamic resource allocationfor load balancing in fog environmentrdquo Wireless Commu-nications amp Mobile Computing vol 2018 no 2 Article ID6421607 15 pages 2018

[92] J Oueis E C Strinati and S Barbarossa ldquoe fog balancingload distribution for small cell cloud computingrdquo in Pro-ceedings of the 2015 IEEE 81st Vehicular Technology Confer-ence (VTC Spring) pp 1ndash6 IEEE Glasgow UK May 2015

[93] K Wang Y Wang Y Sun S Guo and J Wu ldquoGreen in-dustrial internet of things architecture an energy-efficientperspectiverdquo IEEE Communications Magazine vol 54 no 12pp 48ndash54 2016

16 Complexity