research article performance evaluation of a dual...

Research ArticlePerformance Evaluation of a Dual Coverage System forInternet of Things Environments

Omar Said12 and Amr Tolba23

1College of Computers and Information Technology Taif University Taif Saudi Arabia2Mathematics and Computer Science Department Faculty of Science Menoufia University Shebin El Kom 32511 Egypt3Computer Science Department Riyadh Community College King Saud University Riyadh 11437 Saudi Arabia

Correspondence should be addressed to Amr Tolba atolbaksuedusa

Received 10 July 2016 Revised 28 September 2016 Accepted 12 October 2016

Academic Editor Chakchai So-In

Copyright copy 2016 O Said and A TolbaThis is an open access article distributed under the Creative CommonsAttribution Licensewhich permits unrestricted use distribution and reproduction in any medium provided the original work is properly cited

A dual coverage system for Internet ofThings (IoT) environments is introducedThis system is used to connect IoT nodes regardlessof their locations The proposed system has three different architectures which are based on satellites and High Altitude Platforms(HAPs) In case of Internet coverage problems the Internet coveragewill be replacedwith the SatelliteHAPnetwork coverage underspecific restrictions such as loss and delay According to IoT requirements the proposed architectures should includemultiple levelsof satellites or HAPs or a combination of both to cover the global Internet things It was shown that the SatelliteHAPHAPThingsarchitecture provides the largest coverage area A network simulation package NS2 was used to test the performance of theproposed multilevel architectures The results indicated that the HAPHAPThings architecture has the best end-to-end delaypacket loss throughput energy consumption and handover

1 Introduction

During the past three decades the wireless communica-tion field has had notable development New technologyin wireless communication provides users the freedom toconnect to different types of networks Terrestrial and satellitetechnologies provide services toHighData Rate and dynamicnetworks such as in mobile networks Radio Frequency Iden-tification (RFID) networks and Wireless Sensor Networks(WSNs) Terrestrial links provide these services but withhigh complex propagation restrictions Satellite links areused instead of terrestrial links to provide communicationservices for ships planes and TV broadcastingThe satellitesprovide these services to an assigned area on the earth Manyparameters control bandwidth power for the satellite systemssuch as target footprint size traffic control complexity andground station cost Focused satellite transmissions based ontarget goals provide satellite systems with high efficiency dueto interference minimizing and use of efficient spectrums[1ndash10]

HighAltitude Platforms (HAPs) can be constructed usingairships planes or balloons at altitudes of 17ndash22 kilometers

(or higher)They can also deliver a range of services for com-munication and remote sensing with features that are betterthan terrestrial and satellite systems HAPs are competitorswith terrestrial and satellite systems which make themgood candidates for the next generation of communicationsystems In addition HAPs can act as a base station thatcan easily communicate with satellite systems using verytall antennas and can handle the increasing demand ofbroadband wireless access [11ndash13]

Internet of Things (IoT) is a new technology that hasattracted the attention of many network and communicationresearchers in the last few years IoT has many applicationsinmilitary healthcaremarketing and learning [14ndash16] IoT isroughly defined as a computingmodel where various physicalthings communicate with each other using the Internet andprovide information about themselves and their environmentto other passive or active things [17ndash20] IoT has manychallenges associated with diversity in communication nodes(things) large amounts of processing data large numbersof communication nodes (things) routing security energyconsumption for active nodes and things coverage [21] Eachchallenge has important impacts on the performance of IoT

Hindawi Publishing CorporationMobile Information SystemsVolume 2016 Article ID 3464392 20 pageshttpdxdoiorg10115520163464392

2 Mobile Information Systems

systemsThese challengesmay also affect the level of coveragewhich depends on communication technology between IoTnodes It is well known that ground communication betweenthings has a very limited coverage range due to variousimpairments Many existing technologies for IoT nodes havelimited coverage (a fewmeters) [22ndash24]The communicationbetween IoTnodes using satellites orHAPswill providemuchlarger coverage especially with Internet signals which repre-sent amain connectionmedium in IoT systems For examplethe possibility of coverage for large WSNs using HAPs hasbeen proved [25] The radius of this coverage area can beextended to several tens of kilometers In addition the powerof sensors did not need enhancement from external sourcesHAPs provide many advantages over satellites such as lowtransmission power and less delay [26] Because satellites alsohave many advantages such as large coverage areas there arebenefits to constructing and testing multilevel satelliteHAParchitectures Using HAPs in addition to satellites providesefficient IoT systems with satellites at the top of architecturethat link with HAPs used for efficient power transmissionfrom ground or space things For maximum efficiencythese types of architectures may require special hardwaresuch as sensors with parabolic antennas [11 12 27] Usingthese architectures many applications may be constructedto manage IoT system nodes in space and on the earth sur-face These applications can be merged for management inone smartmanagement application which represents the IoTcore idea Different satelliteHAP architectures allow diver-sity in applications which makes IoT systems more flex-ible and scalable Therefore the proposed architecturesare SatelliteHAPHAPThings SatelliteHAPThings andHAPHAPThings

The paper is organized as follows Section 2 presentswork related to HAP and satellite use in IoT In Section 3the research objective and paper contribution are presentedThe dual coverage system and proposed satelliteHAP archi-tectures are introduced in Section 4 Section 5 provides acomparative analysis of the proposed architectures A perfor-mance evaluation of the proposed architectures is providedin Section 6 Finally the conclusion and the future work arepresented in Sections 7 and 8 respectively

2 Related Work

The related work review included work using IoT withsatellite systems WSN with satellite systems and generalInternet satellite systems Few studies were available relatedto IoT and satellite systems De Sanctis et al used a satellitecommunication system to support the IoT [28] This satellitesystem collects data from RFID or sensor systems and sendsdata to actuators This theoretical research did not scale withthe efficiency of the proposed architecture In addition thisresearch did not discuss how the communication betweenthings and satellite would be achieved Limited discussionof IoT and satellites was available on Internet web sites butthese websites presented only general discussion and did notaddress this research area in depth [29]

Many research papers define the relation betweensatellite HAPs andWSN Celandroni et al proposed amodel

for management of disasters remotely This model includesunmanned aerial vehicles (UAVs) equipped with camerasthat are communicated with wirelessly UAVs hover overdisaster areas such as a massive fire or huge traffic accidentand convey information using recording videos and imagesacquired in real time using high technology data sensorsThe extracted information is sent to specialists for evaluationand action determination during disasters This researchconsidered optimal sensor distribution as a challenge Thismodel is not suitable for IoT applications because it used onlyWSNs

The distribution of cells in the HAP coverage area is animportant parameter that affects the capacity ofHAP systemsMany researchers have studied cell distributions and severalproposed dividing HAP coverage areas using 12 to 19 cellsfor capacity enhancement [30 31] Based on this researchYang and Mohammed proposed architectures comprised ofHAPs and a WSN [32] Mitchell et al further developed thisidea and proposed two HAPWSN architectures appropriatefor many vital applications such as monitoring and security[33] The first proposed architecture permits information tobe transmitted directly to the HAP and decreases energyconsumption and complexity It is appropriate for applica-tions that have low data transmission and need large coverageareas The second proposed architecture is comprised ofnodes organized in clusters such that there is no direct com-munication between sensor nodes and HAPs The collecteddata are sent to the cluster head that passed it directly to theHAP Cluster head node selection and energy consumptionare two important challenges in this architecture This archi-tecture is suitable for multimedia applications that transmitlarge numbers of bytes in small intervals

Daniel et al studied the relationship between satellitesand WSN [34] There are many applications using satel-lites and WSNs constructed for emergency communicationsystems and remote area surveillance Many researchershave attempted to solve problems that may occur in satel-liteWSN systems Verma et al evaluated the performanceof a WSNsatellite framework using a network simulationpackage NS2 The simulation calculated many metrics suchas end-to-end delay energy consumption and data gatheringefficiency In addition this research studied the security pro-blem inWSNsatellite systems [35] Another method to eval-uate WSNsatellite performance was constructed by Henautet al Multiband Orthogonal Frequency Division Multiplex-ing (MB-OFDM) was evaluated as a radio interface for newHigh Data Rate (HDR) The results proved that performancewas acceptable for MB-OFDM and WSN applications [36]Raveneau et al proposed an architecture model to inter-connect WSN and satellite technologies This architecture iscalled store-carry-and-forward which is based on the DelayTolerant Networking (DTN) technique In addition a com-parison between new proposed scheduling policy and tradi-tional solutions of DTNwas presented [37] Li et al presentedan algorithm to collect small satellites into one system whereenergy consumption is optimized This algorithm increasedthe network lifetimeThis proposed algorithm is adaptable tonetwork size and the communication mechanism used [38]Amirijoo et al designed a communication server architecture

Mobile Information Systems 3

that is used as a tier between end users and sensor nodeswhere a satellite communication technology is considered amajor link This proposed server design is used to gathersensor data These researchers also presented a dynamicmechanism that adapts to collected data quality [39] Poulakiset al proposed a monitoring application where a Collab-orative Beamforming (CB) mechanism is deployed in theWSNsatellite system without the need for a gateway Ananalysis of the link budget is presented in addition to anexamination of the proposed application under differentnumbers of nodes [40] Shahzad introduced a monitoringsystem that consists of aWSN connected to a satellite systemThis system utilized Google mapping to extract 3D imagesin high resolution This research also utilized an interactiveweb application to decode and sendmessages (gathered data)to a service provider which stores data into a database [41]Mohapatra et al studied location-tracking systems based onWSNs This research used many scenarios and methods toestimate the angle of arrival (AoA) for tracked locations [42]Albagory et al proposed many satelliteHAP architecturesand tested their coverage efficiency but did not discuss meth-ods for applying these architectures on the IoT environmentThe efficiency of these architectures was also not measuredaccording to network metrics such as end-to-end delay andpacket loss ratio [43]

3 Paper Contribution

IoT is an emerging technology that communicates physicalobjects (things) in space in seas and on earth The coverageof these diverse objects is considered a challengeThe Internetis the main transmission medium by which IoT nodes cantransmit their data Despite the large spread of the Internetglobally many objects do not have Internet connectionshowever which presents a problem for IoT scalability Thispaper addresses this problem by introducing a dual coveragesystem to provide IoT nodes with full coverage regardlessof their locations The proposed system has three multilevelarchitectures comprised of four elements satellites HAPsInternet and things The IoT objects will be covered bysatellite or HAPs when unable to access Internet signals Theproposedmultilevel architectures determine the relationshipsbetween satellites HAPs Internet and things In addition aperformance analysis was completed for these architecturesevaluating coverage ability and many network metrics suchas end-to-end delay packet loss ratio throughput energyconsumption and handover

4 Proposed Dual Coverage System

The proposed dual coverage system objective is to guaranteefull coverage for each IoT object regardless of its locationThe proposed systems consist of satellites HAPs Internetand things arranged in three types of multilevel architecturesThese architectures are SatelliteHAPHAPThings Satel-liteHAPThings and HAPHAPThingsThese architecturesmay also include Internet signals Each of the proposedarchitectures has two scenarios for IoT object locations

41 SatelliteHAPHAPThings Architecture The first pro-posed architecture is composed of four layers These layersare satellite HAP HAP and things This architecture is usedto cover isolated areas that are impossible to cover withHAPs alone The first satellite layer satellite can be used incase of communication failure between isolated things andthings covered by the third HAP layer The communicationcost will be decreased by restricting the satellite use Thisarchitecture is similar to the second proposed architecturewhich is described in Section 42 but it has an additionalHAPlayer that may be used in special cases Use of the satellitecan close communication gaps that may result from usingHAPs The location of the first satellite layer is approximatelythousands of kilometers and this layer can be used as aspare communication tool if failure occurs in theHAP secondor third layers The location of the second HAP layer isapproximately located near to 50 kmThe location of the thirdHAP layer is approximately located near to 20 km The IoTobjects may be attached to HAP components or found onthe ground (depending on the needs of the IoT application)Refer to Figures 1 and 2

42 SatelliteHAPThings Architecture The SatelliteHAPThings architecture consists of a satellite backbone HAPthings and Internet connections The IoT objects (passive oractive) should have a direct connection to the HAP Data thatare sent or emitted from passive things should be collectedby the HAP (ie the HAP is considered a sink node for itsregion nodes) Accordingly each of the HAPs should send itscollected data to the satellite backbone The satellite can thenredirect these data to the destination (predetermined objectsin the IoT system) This architecture supposes that there aretwo locations of IoT nodes on the ground and attached toHAPs as shown in Figures 3 and 4 Under this assumptiononly oneHAPnetwork transmits the data which are collectedby other HAPs to the satellite backbone This strategydecreases the communication overhead and system cost Theselection of the HAP used to communicate with the satelliteis an important issue and techniques have been proposed forthis [44 45]The communication between sensors RFIDnet-works and mobile ad hoc networks is possible using Internetor satelliteHAPnetworks as transmissionmediumsThe longdistance between HAP and satellite represents a challengein this architecture because HAP coverage radius is limitedby thing power transmission band and bit rate Thereforethe SatelliteHAPThings architecture is considered a backupcoverage mechanism in case of failing Internet coverageUsing this architecture as a backup coverage system increasesthe overall system cost Hence this architecture should beused as a basic one for many active things such as sensors Inthis case the transmission power will be decreased and theapplications of IoT will became more prevalent

43 HAPHAPThings Architecture Thefirst proposed archi-tecture seems to be costly using a satellite layer As analternative the satellite backbone layer can be replacedwith one or more HAP layers Hence the third proposedarchitecture is composed of three layers The first and secondlayers are HAPs and the third layer is comprised of IoT


Second layer

HAP or HAPs

Third layer

HAPs

Fourth layer

Ground things

Gateway Region 1 Region 2 Region 3 Region 4

Different percentages of Internetcoverage for ground things

Backhaul link

First layer

Satellites

Figure 1 SatelliteHAPHAPThings architecture (scenario 1)

nodes (things) It is well known that HAPs are cheap flexibleand stable comparable with satellites [11 12] HAPs can bereconfigured relocated and repaired in case of failure TheHAP communication system has less transmission delaysand has acceptable links with ground things Furthermorefor mobile users and end users-access providers HAPs havemore efficient communication than satellites These HAPfeatures are adaptable to various IoT application needs How-ever the coverage area of HAPs is small relative to satelliteswhich represents a problem in our proposed architectureOne solution to this problem involves using more HAPs tomaximize the coverage area Additional HAPs increase costbut this cost is still less than a satellite layer In this case inter-HAP links are used to allowHAPs to communicate with eachother

The first HAP layer in this second architecture shouldbe located at approximately 50 km The second HAP layershould also be located at approximately 20 km The thirdlayer is things which may be attached directly to HAPs oron the ground Refer to Figures 5 and 6 The sensitivity ofthing location may represent a challenge in this proposedarchitecture This is because of the definition of IoT whichstates that thing location is a dynamic parameter and requiresthat things should be covered anywhere The second layercollects the data from the third layer comprised of IoT nodes

For example suppose that the third layer has WSN RFIDnetworks and mobile ad hoc networks The communicationbetween these network nodes may be accomplished usingthe second layer In this scenario sending and receivingdata between IoT objects will be achieved using intelligentapplications such as healthcare systems [46ndash48] The firstlayer is used to communicate with second layer HAPs Thecommunication between HAPs in the second layer andbetween the third layer and second layer is simpler due tosatellite replacement [11]

5 Coverage Comparative Study forProposed Architectures

The global coverage for all things is the core objective andcontribution of this paperThe proposed architectures shouldtherefore be compared relative to this objective metric Acomparative study of global earth coverage was done for theproposed architectures The required cellular coverage areadetermines the number of satellites and HAPs that shouldbe used in the target IoT application Suppose that a HAPor satellite is located at an altitude of ℎ km and a minimumelevation angle for covering an area is 119864 So using ℎ and 119864variables the target footprint area can be calculated using


Second layerHAP with

space things

Third layerHAPs with space things

Fourth layerGround things


Different percentages of Internet coverage for ground things

Backhaul link

First layerSatellites withspace things


First layer

Satellites

Second layer

HAPs

Third layer

Ground thingsGateway Region 1 Region 2 Region 3 Region 4


Backhaul link

Figure 3 SatelliteHAPThings architecture (scenario 1)


First layerHAP with

space things

Second layerHAPs with space things

Third layerGround thingsGateway Region 1 Region 2 Region 3 Region 4


Backhaul link


First layer

HAP

Second layer

HAPs

Third layerGround things



Backhaul link

Figure 5 HAPHAPThings architecture (scenario 1)

(1) and (2) below The satelliteHAP geometry is shown inFigure 7

119860 = 21205871199031198902 (1 minus cos (120579)) (1)

120579 = [cosminus1 (119903119890 cos (119864)119903119890 + ℎ )] minus 119864 (2)

The variable 119903119890 is the radius of earth that can be approximatelyevaluated as 6378 km

In this IoT coverage analysis a cellular shape should bedetermined It is supposed to be a hexagonal shape withcircle area 119886 = 1205871199031198882 where 119903119888 is a radius as shown inFigure 8 The actual cell distribution should be treated asa hexagonal shape (not a circular one) due to the circularfootprints which are tessellated with overlapped areas The


First layerHAP with

space things





Backhaul link


SatelliteHAP

h

E

e

re

Figure 7 SatelliteHAP geometry of coverage

cell in the hexagonal view has a radius 119903119888 and its area is givenby

119886ℎ = 3radic32 1199031198882 (3)

Therefore the cell radius which is calculated in (1) has arelation to (3) Hence the resulting area of the cell is givenby

119886ℎ = 3radic31199031198902 (1 minus cos (120579)) (4)

rc

Figure 8 SatelliteHAP coverage footprint of cell

For covering the entire earth with satelliteHAP the numberof satellites and HAPs must be determined This number canbe determined using two relations that are defined in (5) and(6)

119873SH = lceil Area of Earth SurfaceArea of Station Coverage

rceil (5)

119873SH = lceil 41205873radic3 (1 minus cos (120579))rceil (6)

The above analysis is more general and fulfills the IoTapplication needs However the IoT will take a long time tocover the whole earth and become dominant in the worldTherefore we need to clarify how to recover part of the earthFor example if we need to cover a land as a portion of theentire earth space (6) should be decreased by 29 percentThis percentage represents the land ratio relative to the whole


earth Accordingly 119873SH|119871 which represents the coverageland ratio can be calculated using

119873SH1003816100381610038161003816119871 = lceil 1161205873radic3 (1 minus cos (120579))rceil (7)

For a global coverage target a number of satellites and HAPsare needed as determined by (6) and (7) (assuming that thecoverage area equals the cell areas) Hence each of the archi-tectures should be examined to show its coverage feasibilityIn our analysis the geostationary orbit (GEO) is located atan altitude of 36000 km and the low-earth orbit (LEO) islocated at an altitude of 800 kmThese two satellite orbits aremost common orbits With respect to HAPs there are twocommon heightsThe first one is at 20 km representing lowerlayer HAPs and the second is at 50 km representing upperlayer HAPs The results proved that the number of requiredHAPs is much greater than the number of LEO or GEOsatellites This is because the satellites have high altitudesthat provide large coverage areas The coverage of the wholeearth may require approximately one million HAPs with anelevation angle of 55∘ To minimize the number of requiredHAPs we have to increase their coverage areas or reduce theelevation angle Optimization of the required architecture is atarget but this will be addressed in the simulation section Inorder to determine the orbit that fits the required coverageit is mandatory to use satellites in communication betweenIoT nodes For LEO satellites the orbit is not fixed relativeto the earth things In addition the power required for datatransmission is low Accordingly when using LEO satellitesin the proposed architectures the HAPs in the lower layershould communicate with LEO satellites at the visible timesThe LEO satellites should have the ability for switching andtracking However GEO satellites require much more powerfor data transmissions They are also fixed with respect toearth objects The number of required satellites or HAPs atdifferent elevation angles is displayed in Figure 9

6 Simulation and Evaluation

61 Simulation Setup The simulation environment was builtusing the network simulation package NS2 This environ-ment was comprised of five types of networks satelliteHAP WSN RDIF and mobile ad hoc networks There aresix satellites that communicate with each other to create anetwork The data can be redirected from one satellite toanother until it reaches the target satellite Table 1 shows theconfiguration parameters of the satellite network In additionthere are 60 HAPs configured in one network The commu-nication between HAPs may be achieved using inter-HAPstechnology or using a selected satellite [11] Table 2 showsthe configuration parameters of the HAP network Nodes inthe other three networks WSN RFID and mobile ad hocare distributed randomly in the covered areas for HAPs andsatellite networks The percentage of things covered by theInternet is 75 and the remaining 25 of things are coveredby the HAPs and satellite networks The percentage of thingson the ground is 80 with 15 of things in space and 5of things in the sea or underground The simulation of the

Num

ber o

f sta

tions

100

101

102

103

104

105

106

107

Elevation angle (degrees)0 10 20 30 40 50 60

HAP at 20kmHAP at 50km

LEO at 800 kmGEO at 36000 km

Figure 9 SatelliteHAP coverage cell footprint

Table 1 Configuration parameters for satellite simulation

Parameter ValueSatellite type LEOAltitude 800 kmInclination degree 86 (degree)Elevation mask 82 (degree)Uplinkdownlink 15MbsCell size 50 kmPower 1 wattNumber of satellites 4Intersatellite links bandwidth 25MbsIntersatellite links per satellite 6Cross-seam intersatellite links Not foundIntersatellite link delay 78msIntersatellite distance 60 km

IoT environment is flexible since these percentages can bechanged dynamically to get accurate performance results forthe proposed IoT coverage systems Tables 3 4 and 5 containthe configuration parameters of WSN RDIF and mobile adhoc networks respectively

There are four possible simulation scenarios full groundInternet coverage full satelliteHAP network coverage Inter-net over satelliteHAP network coverage and satelliteHAPnetwork with ground Internet coverage The first scenariosupposes that nodes in the IoT environment are covered byground Internet Hence there is no need for satellite andHAP networks as shown in Figure 10 In this scenario eachnode should have Internet connection capability Accord-ingly most of IoT objects are supposed to be active Thisscenario is not considered in the simulation This is because


Table 2 Configuration parameters for HAP simulation

Parameter ValueAltitude 20ndash50 kmBit error rate 10minus6

Elevation mask FlatPower 1 wattReturn channel HAP 30MbsForward channel HAP 60MbsCell size 05 to 10 kmBS max Tx power per link 30 dBmCommon pilot channel Tx power 27 dBmUL load limit 075BS noise figure 5 dBUE max TX power 21 dBmSHO add window 3 dBTraffic bit rate 122 kbpsEbN0UL 5 dBEbN0DL 95 dBSlow fading standard deviation 4 dBDL orthogonality factor 09SHO gain (UL and DL) 1 dBNumber of HAPs 60

Table 3 Configuration parameters for WSN simulation

Parameter ValueFrequency 2400MHzTransmit (TX) data rate 250KbsRF power minus10 dBmReceive (RX) sensitivity minus94 dBmCurrent drain in transmit mode 11mACurrent drain in receive mode 197mABattery 2x 1250mAH 15 VCovered area 1000mtimes 1000mNumber of nodes 1000

Table 4 Configuration parameters for RFID simulation

Parameter ValueData channel frequency 915MHzControl channel frequency 930MHzInterchannel interference NoFading NoSNR based signal reception 10Data rate 2MbpsRadio Rx sensitivity minus91 dBmRx threshold minus81 dBmRFID transmission power minus45 dBmRead range 162 metersSensing range 54 metersInterference range 71 metersNumber of nodes 1200

Table 5 Configuration parameters for mobile ad hoc simulation

Parameter ValuePacket size 1MbNetwork area 500m times 500mTotal number of requests 3200Interval between requestsrsquotransmission 500ms

TTL Random between 4 and 7 msLink availability Between 0 and 1Maximum transmissiondistances 30 to 210m

Maximum node speed 30 kmh to 60 kmhChanging direction probability 119875 = 0Number of nodes 83

using ground Internet makes IoT nodes (things) lack fullcoverage which does not meet our objective The secondscenario supposes that the IoT objects will be covered usinga satelliteHAP network as shown in Figure 11 This scenariodoes not meet the IoT definition because it does not use theInternet as a communication tool between things Thereforethis second scenario is also not considered in the simulationThe third scenario supposes that a part of the IoT nodesis covered by ground Internet and other things are coveredby satellite or HAP as shown in Figure 12 In this scenariothe communication between IoT objects is not unified thatis there are two different communication environmentssatelliteHAPs and ground Internet Accordingly the data canbe transmitted to the HAP it will be directed to the nearestHAP or to the upper satellite layer A gateway should bepresent in this scenario to transmit the data from satellite orHAP networks to the Internet and then to the end user orcentralmanagement systemThis third scenario is consideredin our simulation because it contains two coverage systemssatelliteHAPs and ground Internet that meet our objectiveThe fourth scenario supposes that the IoT nodes are coveredby Internet over satellite orHAPas shown in Figure 13Henceeach node should have an Internet connection using groundInternet or satellite Internet

In our simulation 60 routers are distributed over differentlocations in five countries Each router is connected to anumber of nodes that is determined randomly from 50 to100 nodes (things) The IoT nodes are connected directlyto the router or through a sink node that is used to gatherinformation from its things The relation between routersand HAPs depends on the applied scenario In additionthere are five servers to manage the Internet routers andother networks such as satellite and HAP These servers areused by end users for building interactive IoT applicationsAdditionally there are six LEO-Iridium satellites whichcommunicate with 60 HAPS Each satellite communicateswith 10 HAPs Data is redirected from HAPs in one locationtoHAPs in different locations using the satellitesThe relationbetween HAPs satellites and routers also depends on theexecuted architecture The link bandwidth between satellites


Sink 1

Thing 1Thing 1

Things

Things

Router 1

Sink 1

Thing 1Thing 1

ThingsSink n

Thing 1Thing 1

Router n

Internet things

Figure 10 First simulation scenario full Internet coverage

Thing 1

Thing 2 Thing 3

Thing n

HAP 1 HAP n

Thing 1

Things

middot middot middot

Figure 11 Second simulation scenario full satelliteHAP coverage

is 10Mbs The link bandwidth between HAPs and satellitesis 10Mbs The link bandwidth between HAPs and routers is15Mbs The bandwidth between routers and sinks or thingsrandomly varies between 1 and 2Mbs The propagationdelay of links varies according to the distance between

network components satellites HAPs and routers In oursimulation the propagation delay ranged from 25 to 30msHowever the edge propagation delay was 5ms The routingpaths for transmitting packet streams are determined usingoptimized link state routing protocols [49] in addition toant colony optimization [50] The routing genie is used toadapt the routing process in LEO satellites networks [51 52]In addition in our simulation the things are divided intotwo classes passive and active The passive things use RFIDtechnology to communicate with satellites HAPs or theInternet The active things can send or receive data tofromother things using coverage methods There are four types oftraffic loads video audio image and text MPEG-2 is used asthe compression coding for video streams The compressioncoding for audio streams is PCM The coding of images isJPG Creation of traffic is a randomprocessThe tree topologyis used in the proposed architectures In addition the groundspace and underground nodes are distributed randomly toreflect the real meaning of IoT environment Moreover thetraffic generation model used in this simulation is governedby Poisson distribution Each network component has abuffer to store data packets and redirect them to a predeter-mined destination The buffer size of satellites and HAPs isdetermined using proposed techniques [53 54] The buffersize of Internet components such as routers or servers is arandom value from 100 to 1000 kb For Internet connectionsthe TCP and UDP are used as transport layer protocolsThe selection of the transport layer protocol is achieveddynamically depending on the flow size of packet streams andtypes In case of network starvation (ie the number of lost


Things covered by Internet

Sink 1

Thing 1Thing 1

ThingsThings

Router 1

Sink 1

Thing 1Thing 1

ThingsSink n

Thing 1Thing 1

Router n

Internet things

Thing 1

Thing 2 Thing 3

Thing n

HAP 1 HAP n

Thing 1

Things


25milliseconds15Mbs

Figure 12 Third simulation scenario satelliteHAP and ground Internet coverage

Sink 1

Thing 1Thing 1

Direct connected things

Sink 1

Thing 1Thing 1

Sink n

Thing 1Thing 1

Router 1 Router n

HAP 1 HAP n

Direct connected thingsDirect connected

things

Random speed

Random speed

Random speed

WLAN link

WLAN link

WLAN link

25milliseconds15Mbs

25milliseconds15Mbs

Random speed from 1 to 2Mbs

2Mbs 2Mbs1Mbs



Figure 13 Fourth simulation scenario Internet over satelliteHAP coverage


packets and delay ratio are notably increased) UDP will beused On the contrary in the normal case of the IoT system(ie the IoT metrics have normal values) TCP will be usedThe performance metrics in this simulation are end-to-enddelay packet loss ratio throughput energy consumption andhandover The proposed simulation environment measuresthese performancemetrics for the third and fourth scenariosThe simulation was executed for 1000 minutes For bestresults five simulation iterations were executed and resultsaverages were determined

62 Results and Discussion In this subsection the perfor-mance metrics end-to-end delay packet loss throughputenergy consumption and handover are evaluated and theresults are discussed The performance metrics are measuredfor the three proposed architectures relative to the third andthe fourth simulation scenarios only The first simulationscenario supposes that the Internet is the only coveragetool for IoT objects and neglects satellite and HAP toolsThis scenario is considered a traditional IoT idea that doesnot meet the targeted objective of this paper The secondsimulation scenario also fails to meet the targeted objectivebecause it uses satellite and HAPs only without the Internetand does not meet the typical IoT definition

621 End-To-End Delay The end-to-end delay is consideredan important performance metric due to the large numbersof data packets that may be transmitted through IoT systemsThe end-to-end delay metric is measured from the time apacket is generated to the time it reaches its destination Theend-to-end delay includes the buffering delay that resultsfrom queuing packets at sources and destinationsThe resultsof end-to-end delay analysis are shown in Figures 14 and 15The 119909-axis in the end-to-end graph represents the simulationtime in minutes As stated above the simulation time is 1000minutes The end-to-end delay is calculated by averaging10 delay values (one delay value is extracted every minute)The 119910-axis represents the average end-to-end delay valuesin milliseconds Figure 14 shows the end-to-end delay resultfor the third scenario and Figure 15 shows the end-to-enddelay result for the fourth scenario Result graphs showthat the third simulation scenario has less end-to-end delayvalues than the fourth scenario This can be explained asthe IoT nodes in the third scenario are covered by groundInternet in addition to the satelliteHAP network Howeverthe nodes in the fourth scenario are covered by Internetover the satelliteHAP network which has significant delayscaused by long distances The results shown in Figures 14and 15 indicate that the best average end-to-end delay is forthe HAPHAPThings architecture This is due to the HAPheight (or altitude) being less than the satellite height whichmeans the IoT data can be transmitted with less delay (referto Section 4) In addition at the simulation time point 8the delay is suddenly decreased and smoothly increased inthe next simulation time points This is because the IoTsimulation model includes passive things that are randomlydistributed which means the size of IoT data may increaseor decrease suddenly The hesitation in the plots is also due

0

10

20

30

40

50

60

70

80

90

100

1 10 19 28 37 46 55 64 73 82 91 100

Aver

age d

elay

(ms)

Simulation time (minutes) (10)

SatelliteHAPHapthingsSatelliteHapthingsHAPHapthings

Figure 14 The average end-to-end delay of the third simulationscenario

0

20

40

60

80

100

120

1 10 19 28 37 46 55 64 73 82 91 100

Aver

age d

elay

(ms)


SatelliteHAPHAPthingsSatelliteHAPthingsHAPHAPthings

Figure 15 The average end-to-end delay of the fourth simulationscenario

to bandwidth diversity which is considered a main feature ofIoT environments

622 Packet Loss The packet loss ratio is another importantmetric to assess the delivery performance of data through anIoT environment The packet loss ratio is the percentage oflost packets to the total sent packets within a time intervalunder specific buffer sizes (or window sizes) Packet sequencenumbers are also analyzed for packets which are receivedsuccessfully The average gross number of lost packets withinthe simulation time is measured for the third and thefourth simulation scenarios This metric is measured inenvironments with different power settings encoding typesand bandwidth channels Figure 16 shows the packet loss ratiofor the third simulation scenario and Figure 17 shows the


0

002

004

006

008

01

012

014

016

018

02

1 10 19 28 37 46 55 64 73 82 91 100

Pack

et lo

ss ra

tio


SatelliteHAPHAPthingsSatellite HAPthingsHAPHAPthings

Figure 16 The packet loss ratio of the third simulation scenario

0

01

02

03

04

05

06

07

1 10 19 28 37 46 55 64 73 82 91 100

Pack

et lo

ss ra

tio



Figure 17 The packet loss ratio of the fourth simulation scenario

packet loss ratio for the fourth simulation scenario The 119909-axis represents the packet loss ratio and the 119910-axis representsthe simulation time As stated above in end-to-end delaydiscussion the values of packet ratio are calculated for 10time points For both third and fourth simulation scenariosthe HAPHAPThings architecture has the lowest packetloss ratio and the SatelliteHAPHAPThings architecture hasthe highest packet loss ratio This is due to long distancesbetween the components of the SatelliteHAPHAPThingsarchitecture which in turn reflects on the data delivery per-formanceThe SatelliteHAPThings architecture has a packetloss ratio less than that of the SatelliteHAPHAPThingsarchitecture and larger than that of the HAPHAPThings

architecture Generally the third simulation scenario has alower packet loss ratio than the fourth simulation scenarioThis is explained by the high bit error rate for satellite andHAP comparedwith the bit error rate for the ground InternetIn addition for the third simulation scenario the number ofpackets that are sent through satellite or HAP channels is lessthan the number of packets that are sent through the groundInternet Moreover the number of things that is covered bythe satellite or HAP is less than the number of things that arecovered by ground Internet whichmeans that the packet lossratio of the fourth simulation scenario increases continuitydue to its use of satellite and HAP in the data transmissionprocess This is in contrast to the third simulation scenariowhere most of the IoT data are transmitted by groundInternet Atmost simulation time points the packet loss ratiois stable to some extent However the packet loss ratio at littlepoints of simulation time is extremely high (such as 26 3976 and 80) This is due to the large amount of data that canbe sent at these simulation times from the IoT nodes (largetransmitted data means large packet loss) A sudden increasein transmitted data amounts is due to a sudden increase inthe number of nodes that transmit data which is also animportant feature of IoT environments As stated above thenumber of nodes that can transmit data within interval timesis determined randomly

623 Throughput The throughput of IoT systems can bedefined as the number of bits that are successfully deliveredfrom sources to destinations To determine the efficiencyof IoT systems the throughput metric should be analyzedFigures 18 and 19 show a throughput comparison of thethree proposed architectures relative to the third and thefourth simulation scenarios The results showed that theHAPHAPThings architecture has higher throughput thanthe other two architectures especially when the bit errorrate increases In the satellite systems when the utilizationof the links exceeds specified thresholds the rate of packetloss increases dramatically This explains the throughputsuperiority of the HAPHAPThings architecture in the thirdand the fourth simulation scenarios The distance betweenarchitecture components also plays an important role in thethroughput measurement It is well known that a minimumnumber of nodes in the routing path are a targeted efficiencygoal However sometimes a small number of nodes (withfixed distances between sources and destinations) mean longdistances between intermediate path nodes (routers) whichin turn means high bit error rates like that found in theSatelliteHAPHAPThings and SatelliteHAPThings archi-tectures The sudden increase of throughput value in the firstfive simulation points comes from a sudden increase in thedata sources with low bit error rates and delays After that theplots became stable in the three proposed architectures dueto regular continuity in sending and receiving data tofromthe IoT nodes with high rates The sending and receivingdata in the proposed simulation environments is determinedrandomly with lower and upper limits which mostly pro-vide stability in the transmission rates The passive thingsthroughput value is calculated by the average number of bitssent from their attached RFID tags and received at a specific


0100000020000003000000400000050000006000000700000080000009000000

1000000011000000

1 10 19 28 37 46 55 64 73 82 91 100

Thro

ughp

ut (b

itss

ec) (

10)



Figure 18 The throughput of the third simulation scenario

server (target destination) In Figure 18 minor differencesare shown in the throughput values for the three proposedcoverage architectures In Figure 19 significant differences areshown in the throughput values for the proposed coveragearchitectures which reflect the impact of high packet delaysand losses using the Internet over satelliteHAP networkThethroughput value in the third simulation scenario is higherthan that in the fourth simulation scenario This is explainedby high bit error rates that may be a result of Internet signalsthat are transmitted over satellite or HAP that negativelyaffect the packet loss and the delay metrics In contrastthe ground Internet provides a reliable data transmissionmedium for IoT data Also high throughputs which arefound in the third and the fourth simulation scenario plotsare explained by large numbers of nodes that are simulatedin the IoT environment that provide large and fast datatransmission (in contrast slow transmission decreases thethroughput by consuming large amounts of medium time)

624 Energy Consumption The IoT system is comprised ofenergy-based nodes Energy consumption therefore repre-sents an important factor in the proposed coverage systemEnergy savings for eachnode increase the IoT system lifetimeAs stated above there are three types of networks withenergy-based nodes WSN RFID and mobile ad hoc Totest the energy consumption in the three different networksproposed research techniques were used [55ndash57] In thegraphs of energy consumption the 119909-axis represents 10simulation time points and the 119910-axis represents the valuesof energy consumption For each network each energyconsumption value that is represented on the 119910-axis equalsthe average of node energy consumption within 100 minutesThe number of extracted energy consumption values equals10 (ie 1001000) Figure 20 shows energy consumptionaccording to the third simulation scenario Figures 20(a)

0

1000000

2000000

3000000

4000000

5000000

6000000

7000000

8000000

1 10 19 28 37 46 55 64 73 82 91 100

Thro

ughp

ut (b

itss

ec) (

10)



Figure 19 The throughput of the fourth simulation scenario

20(b) and 20(c) show energy consumption results for WSNRFID and mobile ad hoc network nodes For the threenetworks the HAPHAPThings architecture has the lowestenergy consumption in the most time points The nextlowest energy consumption is in the SatelliteHAPThingsarchitectureThe SatelliteHAPHAPThings architecture hasthe largest energy consumption values It is well knownthat data communication consumes more energy thandata processing or sensing [55] Short-range communi-cation like that in HAPHAPThings architectures savesmore energy than long-range communication like that inSatelliteHAPHAPThings architecture which explains theresults shown in Figure 20 Figure 20 also shows only a fewvalues that contrast with most energy consumption results(such as points 3 4 in WSN points 1 8 and 2 in RFIDand points 1 4 and 7 in the mobile ad hoc network) Thisis explained by the generation of large amounts of data atthese time points which consumes high energy in processingand transmission functions Figure 21 shows the results ofenergy consumption for the three networksWSN RFID andmobile ad hoc according to the fourth simulation scenarioResults shown in Figure 21 are similar to those shown inFigure 20 For the three proposed coverage architecturesthe energy consumption of nodes that are covered usingtools in the third simulation scenario is less than the energyconsumption of nodes that are covered using tools in thefourth simulation scenarioThis is due to the communicationoverhead in addition to the large packet loss ratio which inturn causes retransmission that increases the total number oftransmitted bits

625 Handover Measurement To complete the evaluationof the proposed architectures handover issue should beevaluated The previous Quality of Service (QoS) metricsare evaluated mostly for immobile nodes or nodes that aremoving at slow speed Therefore evaluation of these metricsas regards the mobile nodes completes the performance


1 2 3 4 5 6 7 8 9 10SatelliteHAPHAPthings 5261 5126 475 401 5393 5049 4011 5122 4182 4443SatelliteHAPthings 3985 4363 3498 3297 3431 3691 3923 417 3205 4297HAPHAPthings 3104 3039 386 3918 3086 3531 3896 390 2734 3949

0

100

200

300

400

500

600

Aver

age o

f ene

rgy

cons

umpt

ion

(J)


(a) Energy consumption for WSN nodes third simulation scenario


0

50

100

150

200

250

300

350

400

Aver

age o

f ene

rgy

cons

umpt

ion

(J)


(b) Energy consumption for RFID nodes third simulation scenario


0

20

40

60

80

100

120

140

Aver

age o

f ene

rgy

cons

umpt

ion

(J)


(c) Energy consumption for mobile ad hoc nodes third simulationscenario

Figure 20 Energy consumption in the third simulation scenario

evaluation of the proposed architectures In the proposedarchitectures the flow of signals can be transmitted fromHAP segment to satellites segment or from HAP segment toanother HAP segment Transmission of signals from satellitecoverage area to another satellite coverage area is not con-sidered due to a large area which is covered by one satelliteHence there are two types of handoversHAP-to-Satellite andHAP to HAP Mobile routers DHCP servers wireless LANswireless LANs and RFID networks are main components inthe IoT system The handover process comprises three mainfunctions which are stated as follows information gatheringdecision and execution Information gathering function isused to determine the thresholds of QoS parameters whichare required by the transmitted data The decision is used todetermine whether a handover should be initiated withoutdelay The execution is used to allocate the required QoS forthe transmitted data at the new locationThemobile router isused to achieve the handover process by using mobile server

that represents a home agent (HA) In addition RSVP willbe used as a resource reservation protocol in the proposedarchitectures Moreover mobile IP method is used to addressthe IoT system nodes

To test the handover issue in the proposed architecturesmobile nodes such as mobile phones are used In the sim-ulation environment users are randomly distributed underthe coverage area of HAP or satellite The call generationprocess is governed by Poisson distribution The exponentialdistribution is used to determine the change of call holdingtime The generated traffic is uniformly distributed HAPmovement starts from the center of the coverage area toits end and then it returns to the other end The randomwalk and reflection movements of HAPs are determined in[58] There are three handover metrics which are used todetermine the QoS of mobile calls These metrics are thehandover probability the blocking probability and the dropprobability The HAP speed equals a range between 0 and



0

100

200

300

400

500

600

700

Aver

age o

f ene

rgy

cons

umpt

ion

(J)


(a) Energy consumption for WSN nodes fourth simulation scenario


050

100150200250300350400450500

Aver

age o

f ene

rgy

cons

umpt

ion

(J)


(b) Energy consumption for RFID nodes fourth simulation scenario


020406080

100120140160180

Aver

age o

f ene

rgy

cons

umpt

ion

(J)


(c) Energy consumption for mobile ad hoc nodes fourth simulationscenario

Figure 21 Energy consumption in the fourth simulation scenario

150 kmh the offered traffic equals 2800 the initial positionof HAP is at (0 0 20) km and the user mean arrival call rateequals 10 callshThe speed of LEO satellite equals 11150 kmhThe handover is measured only in the first scenario ofSatelliteHAPHAPThings architectureThe handover in theSatelliteHAPThings and the HAPHAPThings architec-tures is considered a special case form the handover inthe SatelliteHAPHAPThings architecture In addition thethings in the second scenarios of the proposed architecturesare attached directly to the coverage component (space tings)which means that there is no handover occurrence Therelation between satellite and HAP speeds is described insatellite-HAP networks [59]

Figures 22 23 and 24 show the handover probability theblocking probability and the dropping probability respec-tively when the handover occurred for generated calls atinterlayer and intralayer of the SatelliteHAPHAPThingsarchitecture Interlayer means that the calls are transformedfrom HAP to HAP in the same layer Nevertheless intralayermeans that the calls are transformed from one layer to its

upper layerThe 119909-axis represents the speed ofHAPs in kmhThe 119910-axis represents the handover metrics the handoverprobability the blocking probability and the dropping prob-ability The three metrics at the architecture third layer levelhave the lowest values In addition the three metricsrsquo valuesat the second layer level come after the third layer metricsrsquovalues Furthermore the threemetricsrsquo values at the first layerlevel have the largest values This could be explained by thehigh speed of satellite in addition to its high altitude Alsothe threemetricsrsquo values are increasedwith increasing ofHAPor satellite speeds In Figure 22 there are a little number ofspeed points such as 130 at which the blocking probabilityvalue at the first layer is less than that at the second layerThisis due to the little number of calls which may be generated atthis speed point In Figure 23 the handover probability valuesincrease with HAP speed increase without any hesitations orup-normal values In Figure 24 there are hesitations in thefirst and the second layersrsquo plots This could be explained byrandom walk movement of users which makes the droppingprobability plots have more hesitations at the first and the


0

01

02

03

04

05

06

07

08

09

0 10 20 30 40 50 60 70 80 90 100 110 120 130 140 150

Han

dove

r pro

babi

lity

Speed (kmh)

SatelliteHAPHAPthings-third layerSatelliteHAPHAPthings-second layerSatelliteHAPHAPthings-first layer

Figure 22 The handover probability in the SatelliteHAPHAPThings architecture

0

002

004

006

008

01

012

014

016

018

02

022

024

0 10 20 30 40 50 60 70 80 90 100 110 120 130 140 150

Bloc

king

pro

babi

lity

Speed (kmh)


Figure 23 The blocking probability in the SatelliteHAPHAPThings architecture

second layers Moreover it is notable that the three metricsrsquovalues decrease when the speed is decreased This meansthat better channels may be available for the new incomingusersrsquo calls Generally the handover for HAP to HAP evenfor interlayer or intralayer has the lower values as regards theblocking and the dropping probabilities This means that theHAPHAPThings architecture is recommended to guaranteethe required QoS in case of handover

0

005

01

015

02

025

03

035

0 10 20 30 40 50 60 70 80 90 100 110 120 130 140 150

Dro

ppin

g pr

obab

ility

(11

000)

Speed (kmh)


Figure 24 The dropping probability in the SatelliteHAPHAPThings architecture

7 Conclusion

This paper has demonstrated a dual coverage system inwhich IoT objects are covered regardless of their locationsand access to ground Internet This coverage system isimplemented with three main architectures The proposedmultilevel architectures have hierarchical shapes and consistof satellites andor HAPs Two scenarios were examined forthe proposed architectures related to the locations of IoTobjects on the ground or directly attached to satellites orHAPs The research reported in this paper showed that theSatelliteHAPHAPThings architecture provides the largestcoverage area and the HAPHAPThings architecture (witha small number of HAPs) has the lowest coverage area Asimulation environment was constructed using a networksimulation package NS2 to test the performance of theproposed architectures under two scenarios The simulationresults showed that the HAPHAPThings architecture hasthe lowest end-to-end delay packet loss ratio and nodesenergy consumption in addition to the largest throughputand smooth handover when compared to the other proposedarchitectures The SatelliteHAPGround Internet scenario isrecommended rather than the Internet over SatelliteHAPnetwork scenario assuming that the number of HAPs is aconsidered parameter in the coverage architecture design

8 Future Work

In the future work different codings of multimedia suchas MPEG-4 should be transmitted through the proposedarchitectures and the results should be discussed In additionin order to get the nearest spec of IoT environment thecomplexity of simulation should be increased Furthermore


study of data flow through the components of each layer inthe proposed architectures should be accomplished

Competing Interests

The authors declare that there is no conflict of interestsregarding the publication of this paper

Acknowledgments

The authors are thankful to the Deanship of ScientificResearch King Saud University Riyadh Saudi Arabia forfunding this research work

References

[1] Y Wang M Wilkerson and X Yu ldquoHybrid sensor deploy-ment for surveillance and target detection in wireless sensornetworksrdquo in Proceedings of the 7th International WirelessCommunications and Mobile Computing Conference (IWCMCrsquo11) pp 326ndash330 Istanbul Turkey July 2011

[2] R Alageswaran R Usha R Gayathridevi and G KiruthikaldquoDesign and implementation of dynamic sink node placementusing particle swarm optimization for life time maximizationof WSN applicationsrdquo in Proceedings of the IEEE InternationalConference on Advances in Engineering Science and Manage-ment (ICAESM rsquo12) pp 552ndash555Nagapattin IndiaMarch 2012

[3] P Nie J Salminen L Andrey and A Yla-Jaaski ldquoSmart triggerfor ultralow power and time critical WSN applicationsrdquo inProceedings of the IEEE International Conference on GreenComputing and Communications (GreenCom rsquo12) pp 163ndash170Besancon France November 2012

[4] W Yu and X Qian ldquoDesign of 3KW wind and solar hybridindependent power supply system for 3G base stationrdquo inProceedings of the 2nd International Symposium on KnowledgeAcquisition and Modeling (KAM rsquo09) pp 289ndash292 WuhanChina December 2009

[5] H Suo J Wan C Zou and J Liu ldquoSecurity in the internet ofthings a reviewrdquo in Proceedings of the International Conferenceon Computer Science and Electronics Engineering (ICCSEE rsquo12)pp 648ndash651 IEEE Guangzhou China March 2012

[6] W Zhaofeng H Guyu Y Seyedi and J Fenglin ldquoA simple real-time handovermanagement in themobile satellite communica-tion networksrdquo in Proceedings of the 17th Asia-Pacific NetworkOperations and Management Symposium (APNOMS rsquo15) pp175ndash179 Busan South Korea August 2015

[7] S Berrezzoug F T Bendimerad and A Boudjemai ldquoCommu-nication satellite link budget optimization using gravitationalsearch algorithmrdquo in Proceedings of the 3rd International Con-ference onControl Engineeringamp Information Technology (CEITrsquo15) pp 1ndash7 IEEE Tlemcen Algeria May 2015

[8] J Li G-Q Ye J Zhang T-J Zhang and L-J Ke ldquoA routingalgorithm satisfied ground station distribution constraint forsatellite constellation networkrdquo in Proceedings of the Science andInformation Conference (SAI rsquo15) pp 997ndash1002 London UKJuly 2015

[9] S Manzari S Caizzone C Rubini and G Marrocco ldquoFeasi-bility of wireless temperature sensing by passive UHF-RFIDtags in ground satellite test bedsrdquo in Proceedings of the 2ndInternational IEEEConference onWireless for Space andExtreme

Environments (WiSEE rsquo14) pp 1ndash6 IEEE Noordwijk TheNetherlands October 2014

[10] N Celandroni E Ferro A Gotta et al ldquoA survey of architec-tures and scenarios in satellite-based wireless sensor networkssystem design aspectsrdquo International Journal of Satellite Com-munications and Networking vol 31 no 1 pp 1ndash38 2013

[11] P Pace G Aloi F De Rango E Natalizio A Molinaro and SMarano ldquoAn integrated Satellite-HAP-Terrestrial system archi-tecture resources allocation and traffic management issuesrdquoin Proceedings of the 2004 IEEE 59th Vehicular TechnologyConference (VTC rsquo04) pp 2872ndash2875 Milan Italy May 2004

[12] S H Alsamhi and N S Rajput ldquoHAP antenna radiation patt-ern for providing coverage and service characteristicsrdquo inProceedings of the 3rd International Conference on Advances inComputing Communications and Informatics (ICACCI rsquo14) pp1434ndash1439 September 2014

[13] Y Albagory and O Said ldquoPerformance enhancement of high-altitude platforms wireless sensor networks using concentriccircular arraysrdquo AEUmdashInternational Journal of Electronics andCommunications vol 69 no 1 pp 382ndash388 2015

[14] M Nitti L Atzori and I P Cvijikj ldquoNetwork navigability inthe social Internet of Thingsrdquo in Proceedings of the IEEE WorldForum on Internet of Things (WF-IoT rsquo14) pp 405ndash410 IEEESeoul Republic of Korea March 2014

[15] X Cheng and G Dang ldquoThe P2P communication technologyresearch based on internet of thingsrdquo in Proceedings of theIEEEWorkshop on Advanced Research and Technology in Indus-try Applications (WARTIA rsquo14) pp 178ndash180 Ottawa CanadaSeptember 2014

[16] L Zhang X Wang C Wang and X Gu ldquoThe application ofstolen radioactive source tracking system based on internet ofthings technologyrdquo in Proceedings of the 3rd International Con-ference on Measuring Technology and Mechatronics Automation(ICMTMA rsquo11) pp 696ndash698 Shanghai China January 2011

[17] D Singh G Tripathi and A J Jara ldquoA survey of internet-of-things future vision architecture challenges and servicesrdquo inProceedings of the IEEEWorld Forum on Internet ofThings (WF-IoT rsquo14) pp 287ndash292 Seoul Korea March 2014

[18] L Catarinucci D de Donno L Mainetti et al ldquoAn IoT-awarearchitecture for smart healthcare systemsrdquo IEEE Internet ofThings Journal vol 2 no 6 pp 515ndash526 2015

[19] Q Yongrui Q Z Sheng N J G Falkner S Dustdar H Wangand A V Vasilakos ldquoWhen things matter a survey on data-centric internet of thingsrdquo Journal of Network and ComputerApplications vol 64 pp 137ndash153 2016

[20] B Horan M Gardner and J Scott ldquoMiRTLE a mixed realityteaching amp learning environmentrdquo Technical Report of SunMicrosystems Laboratories University of Essex ColchesterUK 2009

[21] N Bari G Mani and S Berkovich ldquoInternet of things as amethodological conceptrdquo in Proceedings of the 4th InternationalConference on Computing for Geospatial Research and Applica-tion (COMGeo rsquo13) pp 48ndash55 San Jose Calif USA July 2013

[22] I Ishaq J Hoebeke I Moerman and P Demeester ldquoInternetof things virtual networks bringing network virtualization toresource-constrained devicesrdquo in Proceedings of the IEEE Inter-national Conference on Green Computing and Communications(GreenCom rsquo12) pp 293ndash300 Besancon France November2012

[23] J Dong J Han J Liu and H Xu ldquoThe shallow analysis of theenlightenment of cloud computing to distance educationrdquo


in Proceedings of the International Conference on E-HealthNetworking Digital Ecosystems and Technologies (EDT rsquo10) pp301ndash303 IEEE Shenzhen China April 2010

[24] O Said and M Masud ldquoTowards internet of things survey andfuture visionrdquo International Journal of Computer Networks vol5 no 1 pp 1ndash17 2013

[25] F Yin Z Li and H Wang ldquoEnergy-efficient data collection inmultiple mobile gateways WSN-MCN convergence systemrdquo inProceedings of the 2013 IEEE 10th Consumer Communicationsand Networking Conference (CCNC rsquo13) pp 271ndash276 Las VegasNev USA January 2013

[26] M F Urso M Mondin E Falletti F Sellone and S ArnonldquoSelf organizing WSN collaborative beamforming for HAPcommunicationsrdquo in Proceedings of the IEEE GLOBECOMWorkshops pp 1ndash5 IEEE New Orleans La USA December2008

[27] M2M and IOT via Satellite 6th Edition Market ResearchReportndash249859 Northern Sky Research LLC httpwwwgii-researchcomreportns249859-scada-m2m-via-satellite-3rd-edi-tionhtml

[28] M De Sanctis E Cianca G Araniti I Bisio and R Pra-sad ldquoSatellite communications supporting internet of remotethingsrdquo IEEE Internet ofThings Journal vol 3 no 1 pp 113ndash1232016

[29] httpwwwthurayacomcontentcan-internet-things-iot-sur-vive-without-satellite

[30] M Quaritsch K Kruggl D Wischounig-Strucl S Bhat-tacharya M Shah and B Rinner ldquoNetworked UAVs as aerialsensor network for disaster management applicationsrdquo Elek-trotechnik und Informationstechnik vol 127 no 3 pp 56ndash632010

[31] Z Yang and A Mohammed ldquoHigh altitude platforms forwireless sensor network applicationsrdquo in Proceedings of the IEEEInternational Symposium on Wireless Communication Systems(ISWCS rsquo08) pp 613ndash617 Reykjavik Iceland October 2008

[32] Z Yang andAMohammed ldquoA study ofmultiple access schemesfor wireless sensor network applications via high altitude sys-temsrdquo in Proceedings of IEEE 69th International Conference onVehicular Technology Conference (VTC rsquo09) pp 1ndash5 BarcelonaSpain April 2009

[33] P D Mitchell J Qiu H Li and D Grace ldquoUse of aerialplatforms for energy efficient medium access control in wirelesssensor networksrdquo Computer Communications vol 33 no 4 pp500ndash512 2010

[34] K Daniel S Rohde N Goddemeier and CWietfeld ldquoChannelaware mobility for self organizing wireless sensor swarms basedon low altitude platformsrdquo inProceedings of the 7th InternationalSymposium on Wireless Communication Systems (ISWCS rsquo10)pp 145ndash149 York UK September 2010

[35] S Verma P Pillai and Y F Hu ldquoPerformance analysis of dataaggregation and security inWSN-satellite integrated networksrdquoin Proceedings of the IEEE 24th Annual International SymposiumonPersonal Indoor andMobile Radio Communications (PIMRCrsquo13) pp 3297ndash3301 London UK September 2013

[36] J Henaut D Dragomirescu F Perget and R Plana ldquoValidationof the MB-OFDM modulation for High Data Rate WSN forsatellite ground testingrdquo in Proceedings of the 5th InternationalConference on Systems (ICONS rsquo10) pp 41ndash46 MenuiresFrance April 2010

[37] P Raveneau E Chaput R Dhaou E Dubois P Gelard and A-L Beylot ldquoCarreau CARrier REsource access for mUle DTN

applied to hybrid WSNsatellite systemrdquo in Proceedings of the2013 IEEE 78th Vehicular Technology Conference (VTCrsquo 13) LasVegas Nev USA September 2013

[38] W Li T Arslan J Han et al ldquoEnergy efficiency enhancementin satellite basedWSN through collaboration and self-organizedmobilityrdquo in Proceedings of the IEEE Aerospace Conference pp1ndash8 Big Sky Mont USA March 2009

[39] M Amirijoo S H Son and J Hansson ldquoQoD adaptation forachieving lifetime predictability ofWSN nodes communicatingover satellite linksrdquo in Proceedings of the 4th InternationalConference on Networked Sensing Systems (INSS rsquo07) pp 19ndash26Braunschweig Germany June 2007

[40] M I Poulakis S Vassaki and A D Panagopoulos ldquoSatellite-based wireless sensor networks radio communication linkdesignrdquo in Proceedings of the 7th European Conference onAnten-nas and Propagation (EuCAP rsquo13) pp 2620ndash2624 GothenburgSweden April 2013

[41] F Shahzad ldquoSatellite monitoring of Wireless Sensor Networks(WSNs)rdquo Procedia Computer Science vol 21 pp 479ndash484 2013

[42] S Mohapatra V SurendraSai and C Tripathy ldquoA comparativeview of AoA estimation inWSN positioningrdquo inComputationalIntelligence in Data MiningmdashVolume 3 Proceedings of theInternational Conference on CIDM 20-21 December 2014 vol33 of Smart Innovation Systems and Technologies pp 123ndash133Springer Berlin Germany 2014

[43] YAlbagory FAl Raddady SAljahdali andO Said ldquoInnovativelarge scale wireless sensor network architecture using satellitesand high-altitude platformsrdquo International Journal of Wirelessand Microwave Technologies vol 4 no 2 pp 12ndash19 2014

[44] Z Yang and A Mohammed Wireless Sensor NetworksApplications via High Altitude Systems Emerging Commu-nications for Wireless Sensor Networks InTech 2011 httpwwwintechopencombooksemerging-communications-for-wireless-sensor-networkswireless-sensor-networks-applications-via-high-altitude-systems

[45] M Hamdi L Franck and X Lagrange ldquoNovel cluster main-tenance protocol for efficient satellite integration in MANETsrdquoin Proceedings of the 29th AIAA International CommunicationsSatellite Systems Conference (ICSSC rsquo11) pp 2ndash10 Nara JapanDecember 2011

[46] H-B Li T Takahashi M Toyoda N Katayama YMori and RKohno ldquoAn experimental system enablingWBANdata deliveryvia satellite communication linksrdquo in Proceedings of the IEEEInternational Symposium on Wireless Communication Systems(ISWCS rsquo08) pp 354ndash358 Reykjavik Iceland October 2008

[47] I Sachpazidis D Rizou andW Menary ldquoSatellite based healthnetwork in Peru and Brazilrdquo in Proceedings of the InternationalConference on Information Technology and Applications inBiomedicine (ITAB rsquo08) pp 309ndash314 Shenzhen China May2008

[48] H Ben Elhadj J Elias L Chaari and L Kamoun ldquoA prioritybased cross layer routing protocol for healthcare applicationsrdquoAd Hoc Networks vol 42 pp 1ndash18 2016

[49] C Rodriguez L Franck C Baudoin and A Beylot ldquoOLSR-Ha satellite-terrestrial hybrid broadcasting for OLSR signalingrdquoin Personal Satellite Services Third International ICST Confer-ence PSATS 2011 Malaga Spain February 17-18 2011 RevisedSelected Papers vol 71 of Lecture Notes of the Institute forComputer Sciences Social Informatics and TelecommunicationsEngineering pp 143ndash150 Springer Berlin Germany 2011

[50] H Wang M Xu R Wang and Y Li ldquoScheduling earth observ-ing satellites with hybrid ant colony optimization algorithmrdquo


in Proceedings of 2009 International Conference on ArtificialIntelligence and Computational Intelligence (AICI rsquo09) pp 245ndash249 Shanghai China November 2009

[51] L Audah Z Sun and H Cruickshank ldquoEnd-to-end QoS eval-uation of IP-Diffserv network over LEO satellite constellationrdquoLecture Notes of the Institute for Computer Sciences SocialInformatics and Telecommunications Engineering vol 43 pp99ndash113 2010

[52] K Fall and K VaradhanTheNSManual University CaliforniaBerkeley Calif USA 2008

[53] M Emmelmann ldquoEffects of advertised receive buffer size andtimer granularity onTCPperformance over erroneous links in aLEO satellite networkrdquo in Proceedings of the IEEE Conference onGlobal Telecommunications (GLOBECOM rsquo02) vol 3 pp 2955ndash2958 Taipei Taiwan November 2002

[54] M Knapek J Horwath F Moll B Epple and N CourvilleldquoOptical high-capacity satellite downlinks via high-altitudeplatform relaysrdquo inProceedings of SPIE-The International Societyfor Optical Engineering Free-Space Laser Communications VIIvol SPIE 6709 September 2007

[55] H Zhou D Luo Y Gao and D Zuo ldquoModeling of node energyconsumption for wireless sensor networksrdquo Wireless SensorNetwork vol 3 no 1 pp 18ndash23 2011

[56] X Yan and X Liu ldquoEvaluating the energy consumption ofthe RFID tag collision resolution protocolsrdquoTelecommunicationSystems vol 52 no 4 pp 2561ndash2568 2013

[57] HXiaoDM Ibrahim andBChristianson ldquoEnergy consump-tion in mobile ad hoc networksrdquo in Proceedings of the IEEEWireless Communications and Networking Conference (WCNCrsquo14) pp 2599ndash2604 IEEE Istanbul Turkey April 2014

[58] SWang L Sun F Xiao X Ye and RWang ldquoA newTCP designfor satellite-HAP networksrdquo Communications in Computer andInformation Science vol 334 pp 467ndash477 2013

[59] F Dong H Li X Gong Q Liu and J Wang ldquoEnergy-efficient transmissions for remote wireless sensor networks anintegrated HAPsatellite architecture for emergency scenariosrdquoSensors vol 15 no 9 pp 22266ndash22290 2015

Submit your manuscripts athttpwwwhindawicom

Computer Games Technology

International Journal of

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014


Distributed Sensor Networks


Advances in

FuzzySystems

Hindawi Publishing Corporationhttpwwwhindawicom

Volume 2014


ReconfigurableComputing

Hindawi Publishing Corporation httpwwwhindawicom Volume 2014


Applied Computational Intelligence and Soft Computing

thinspAdvancesthinspinthinsp

Artificial Intelligence

HindawithinspPublishingthinspCorporationhttpwwwhindawicom Volumethinsp2014

Advances inSoftware EngineeringHindawi Publishing Corporationhttpwwwhindawicom Volume 2014


Electrical and Computer Engineering

Journal of

Journal of

Computer Networks and Communications


Hindawi Publishing Corporation

httpwwwhindawicom Volume 2014

Advances in

Multimedia


Biomedical Imaging


ArtificialNeural Systems

Advances in


RoboticsJournal of



Computational Intelligence and Neuroscience

Industrial EngineeringJournal of


Modelling amp Simulation in EngineeringHindawi Publishing Corporation httpwwwhindawicom Volume 2014

The Scientific World JournalHindawi Publishing Corporation httpwwwhindawicom Volume 2014


Human-ComputerInteraction

Advances in

Computer EngineeringAdvances in



systemsThese challengesmay also affect the level of coveragewhich depends on communication technology between IoTnodes It is well known that ground communication betweenthings has a very limited coverage range due to variousimpairments Many existing technologies for IoT nodes havelimited coverage (a fewmeters) [22ndash24]The communicationbetween IoTnodes using satellites orHAPswill providemuchlarger coverage especially with Internet signals which repre-sent amain connectionmedium in IoT systems For examplethe possibility of coverage for large WSNs using HAPs hasbeen proved [25] The radius of this coverage area can beextended to several tens of kilometers In addition the powerof sensors did not need enhancement from external sourcesHAPs provide many advantages over satellites such as lowtransmission power and less delay [26] Because satellites alsohave many advantages such as large coverage areas there arebenefits to constructing and testing multilevel satelliteHAParchitectures Using HAPs in addition to satellites providesefficient IoT systems with satellites at the top of architecturethat link with HAPs used for efficient power transmissionfrom ground or space things For maximum efficiencythese types of architectures may require special hardwaresuch as sensors with parabolic antennas [11 12 27] Usingthese architectures many applications may be constructedto manage IoT system nodes in space and on the earth sur-face These applications can be merged for management inone smartmanagement application which represents the IoTcore idea Different satelliteHAP architectures allow diver-sity in applications which makes IoT systems more flex-ible and scalable Therefore the proposed architecturesare SatelliteHAPHAPThings SatelliteHAPThings andHAPHAPThings

The paper is organized as follows Section 2 presentswork related to HAP and satellite use in IoT In Section 3the research objective and paper contribution are presentedThe dual coverage system and proposed satelliteHAP archi-tectures are introduced in Section 4 Section 5 provides acomparative analysis of the proposed architectures A perfor-mance evaluation of the proposed architectures is providedin Section 6 Finally the conclusion and the future work arepresented in Sections 7 and 8 respectively

2 Related Work

The related work review included work using IoT withsatellite systems WSN with satellite systems and generalInternet satellite systems Few studies were available relatedto IoT and satellite systems De Sanctis et al used a satellitecommunication system to support the IoT [28] This satellitesystem collects data from RFID or sensor systems and sendsdata to actuators This theoretical research did not scale withthe efficiency of the proposed architecture In addition thisresearch did not discuss how the communication betweenthings and satellite would be achieved Limited discussionof IoT and satellites was available on Internet web sites butthese websites presented only general discussion and did notaddress this research area in depth [29]

Many research papers define the relation betweensatellite HAPs andWSN Celandroni et al proposed amodel

for management of disasters remotely This model includesunmanned aerial vehicles (UAVs) equipped with camerasthat are communicated with wirelessly UAVs hover overdisaster areas such as a massive fire or huge traffic accidentand convey information using recording videos and imagesacquired in real time using high technology data sensorsThe extracted information is sent to specialists for evaluationand action determination during disasters This researchconsidered optimal sensor distribution as a challenge Thismodel is not suitable for IoT applications because it used onlyWSNs

The distribution of cells in the HAP coverage area is animportant parameter that affects the capacity ofHAP systemsMany researchers have studied cell distributions and severalproposed dividing HAP coverage areas using 12 to 19 cellsfor capacity enhancement [30 31] Based on this researchYang and Mohammed proposed architectures comprised ofHAPs and a WSN [32] Mitchell et al further developed thisidea and proposed two HAPWSN architectures appropriatefor many vital applications such as monitoring and security[33] The first proposed architecture permits information tobe transmitted directly to the HAP and decreases energyconsumption and complexity It is appropriate for applica-tions that have low data transmission and need large coverageareas The second proposed architecture is comprised ofnodes organized in clusters such that there is no direct com-munication between sensor nodes and HAPs The collecteddata are sent to the cluster head that passed it directly to theHAP Cluster head node selection and energy consumptionare two important challenges in this architecture This archi-tecture is suitable for multimedia applications that transmitlarge numbers of bytes in small intervals

Daniel et al studied the relationship between satellitesand WSN [34] There are many applications using satel-lites and WSNs constructed for emergency communicationsystems and remote area surveillance Many researchershave attempted to solve problems that may occur in satel-liteWSN systems Verma et al evaluated the performanceof a WSNsatellite framework using a network simulationpackage NS2 The simulation calculated many metrics suchas end-to-end delay energy consumption and data gatheringefficiency In addition this research studied the security pro-blem inWSNsatellite systems [35] Another method to eval-uate WSNsatellite performance was constructed by Henautet al Multiband Orthogonal Frequency Division Multiplex-ing (MB-OFDM) was evaluated as a radio interface for newHigh Data Rate (HDR) The results proved that performancewas acceptable for MB-OFDM and WSN applications [36]Raveneau et al proposed an architecture model to inter-connect WSN and satellite technologies This architecture iscalled store-carry-and-forward which is based on the DelayTolerant Networking (DTN) technique In addition a com-parison between new proposed scheduling policy and tradi-tional solutions of DTNwas presented [37] Li et al presentedan algorithm to collect small satellites into one system whereenergy consumption is optimized This algorithm increasedthe network lifetimeThis proposed algorithm is adaptable tonetwork size and the communication mechanism used [38]Amirijoo et al designed a communication server architecture











Second layer

HAP or HAPs

Third layer

HAPs

Fourth layer

Ground things



Backhaul link

First layer

Satellites









space things





Backhaul link



First layer

Satellites

Second layer

HAPs

Third layer



Backhaul link



First layerHAP with

space things




Backhaul link


First layer

HAP

Second layer

HAPs




Backhaul link



119860 = 21205871199031198902 (1 minus cos (120579)) (1)





First layerHAP with

space things





Backhaul link


SatelliteHAP

h

E

e

re



119886ℎ = 3radic32 1199031198882 (3)


119886ℎ = 3radic31199031198902 (1 minus cos (120579)) (4)

rc




rceil (5)









Num

ber o

f sta

tions

100

101

102

103

104

105

106

107
























Sink 1

Thing 1Thing 1

Things

Things

Router 1

Sink 1

Thing 1Thing 1

ThingsSink n

Thing 1Thing 1

Router n

Internet things


Thing 1

Thing 2 Thing 3

Thing n

HAP 1 HAP n

Thing 1

Things







Sink 1

Thing 1Thing 1

ThingsThings

Router 1

Sink 1

Thing 1Thing 1

ThingsSink n

Thing 1Thing 1

Router n

Internet things

Thing 1

Thing 2 Thing 3

Thing n

HAP 1 HAP n

Thing 1

Things


25milliseconds15Mbs


Sink 1

Thing 1Thing 1


Sink 1

Thing 1Thing 1

Sink n

Thing 1Thing 1

Router 1 Router n

HAP 1 HAP n


things

Random speed

Random speed

Random speed

WLAN link

WLAN link

WLAN link

25milliseconds15Mbs

25milliseconds15Mbs


2Mbs 2Mbs1Mbs








0

10

20

30

40

50

60

70

80

90

100

1 10 19 28 37 46 55 64 73 82 91 100

Aver

age d

elay

(ms)




0

20

40

60

80

100

120

1 10 19 28 37 46 55 64 73 82 91 100

Aver

age d

elay

(ms)







0

002

004

006

008

01

012

014

016

018

02

1 10 19 28 37 46 55 64 73 82 91 100

Pack

et lo

ss ra

tio




0

01

02

03

04

05

06

07

1 10 19 28 37 46 55 64 73 82 91 100

Pack

et lo

ss ra

tio








0100000020000003000000400000050000006000000700000080000009000000

1000000011000000

1 10 19 28 37 46 55 64 73 82 91 100

Thro

ughp

ut (b

itss

ec) (

10)






0

1000000

2000000

3000000

4000000

5000000

6000000

7000000

8000000

1 10 19 28 37 46 55 64 73 82 91 100

Thro

ughp

ut (b

itss

ec) (

10)








0

100

200

300

400

500

600

Aver

age o

f ene

rgy

cons

umpt

ion

(J)




0

50

100

150

200

250

300

350

400

Aver

age o

f ene

rgy

cons

umpt

ion

(J)




0

20

40

60

80

100

120

140

Aver

age o

f ene

rgy

cons

umpt

ion

(J)









0

100

200

300

400

500

600

700

Aver

age o

f ene

rgy

cons

umpt

ion

(J)




050

100150200250300350400450500

Aver

age o

f ene

rgy

cons

umpt

ion

(J)




020406080

100120140160180

Aver

age o

f ene

rgy

cons

umpt

ion

(J)








0

01

02

03

04

05

06

07

08

09

0 10 20 30 40 50 60 70 80 90 100 110 120 130 140 150

Han

dove

r pro

babi

lity

Speed (kmh)



0

002

004

006

008

01

012

014

016

018

02

022

024

0 10 20 30 40 50 60 70 80 90 100 110 120 130 140 150

Bloc

king

pro

babi

lity

Speed (kmh)




0

005

01

015

02

025

03

035

0 10 20 30 40 50 60 70 80 90 100 110 120 130 140 150

Dro

ppin

g pr

obab

ility

(11

000)

Speed (kmh)



7 Conclusion


8 Future Work




Competing Interests


Acknowledgments


References









































































Advances in

FuzzySystems


Volume 2014












Journal of

Journal of





Advances in

Multimedia


Biomedical Imaging



Advances in


RoboticsJournal of










Advances in













Second layer

HAP or HAPs

Third layer

HAPs

Fourth layer

Ground things



Backhaul link

First layer

Satellites









space things





Backhaul link



First layer

Satellites

Second layer

HAPs

Third layer



Backhaul link



First layerHAP with

space things




Backhaul link


First layer

HAP

Second layer

HAPs




Backhaul link



119860 = 21205871199031198902 (1 minus cos (120579)) (1)





First layerHAP with

space things





Backhaul link


SatelliteHAP

h

E

e

re



119886ℎ = 3radic32 1199031198882 (3)


119886ℎ = 3radic31199031198902 (1 minus cos (120579)) (4)

rc




rceil (5)









Num

ber o

f sta

tions

100

101

102

103

104

105

106

107
























Sink 1

Thing 1Thing 1

Things

Things

Router 1

Sink 1

Thing 1Thing 1

ThingsSink n

Thing 1Thing 1

Router n

Internet things


Thing 1

Thing 2 Thing 3

Thing n

HAP 1 HAP n

Thing 1

Things







Sink 1

Thing 1Thing 1

ThingsThings

Router 1

Sink 1

Thing 1Thing 1

ThingsSink n

Thing 1Thing 1

Router n

Internet things

Thing 1

Thing 2 Thing 3

Thing n

HAP 1 HAP n

Thing 1

Things


25milliseconds15Mbs


Sink 1

Thing 1Thing 1


Sink 1

Thing 1Thing 1

Sink n

Thing 1Thing 1

Router 1 Router n

HAP 1 HAP n


things

Random speed

Random speed

Random speed

WLAN link

WLAN link

WLAN link

25milliseconds15Mbs

25milliseconds15Mbs


2Mbs 2Mbs1Mbs








0

10

20

30

40

50

60

70

80

90

100

1 10 19 28 37 46 55 64 73 82 91 100

Aver

age d

elay

(ms)




0

20

40

60

80

100

120

1 10 19 28 37 46 55 64 73 82 91 100

Aver

age d

elay

(ms)







0

002

004

006

008

01

012

014

016

018

02

1 10 19 28 37 46 55 64 73 82 91 100

Pack

et lo

ss ra

tio




0

01

02

03

04

05

06

07

1 10 19 28 37 46 55 64 73 82 91 100

Pack

et lo

ss ra

tio








0100000020000003000000400000050000006000000700000080000009000000

1000000011000000

1 10 19 28 37 46 55 64 73 82 91 100

Thro

ughp

ut (b

itss

ec) (

10)






0

1000000

2000000

3000000

4000000

5000000

6000000

7000000

8000000

1 10 19 28 37 46 55 64 73 82 91 100

Thro

ughp

ut (b

itss

ec) (

10)








0

100

200

300

400

500

600

Aver

age o

f ene

rgy

cons

umpt

ion

(J)




0

50

100

150

200

250

300

350

400

Aver

age o

f ene

rgy

cons

umpt

ion

(J)




0

20

40

60

80

100

120

140

Aver

age o

f ene

rgy

cons

umpt

ion

(J)









0

100

200

300

400

500

600

700

Aver

age o

f ene

rgy

cons

umpt

ion

(J)




050

100150200250300350400450500

Aver

age o

f ene

rgy

cons

umpt

ion

(J)




020406080

100120140160180

Aver

age o

f ene

rgy

cons

umpt

ion

(J)








0

01

02

03

04

05

06

07

08

09

0 10 20 30 40 50 60 70 80 90 100 110 120 130 140 150

Han

dove

r pro

babi

lity

Speed (kmh)



0

002

004

006

008

01

012

014

016

018

02

022

024

0 10 20 30 40 50 60 70 80 90 100 110 120 130 140 150

Bloc

king

pro

babi

lity

Speed (kmh)




0

005

01

015

02

025

03

035

0 10 20 30 40 50 60 70 80 90 100 110 120 130 140 150

Dro

ppin

g pr

obab

ility

(11

000)

Speed (kmh)



7 Conclusion


8 Future Work




Competing Interests


Acknowledgments


References









































































Advances in

FuzzySystems


Volume 2014












Journal of

Journal of





Advances in

Multimedia


Biomedical Imaging



Advances in


RoboticsJournal of










Advances in





space things





Backhaul link



First layer

Satellites

Second layer

HAPs

Third layer



Backhaul link



First layerHAP with

space things




Backhaul link


First layer

HAP

Second layer

HAPs




Backhaul link



119860 = 21205871199031198902 (1 minus cos (120579)) (1)





First layerHAP with

space things





Backhaul link


SatelliteHAP

h

E

e

re



119886ℎ = 3radic32 1199031198882 (3)


119886ℎ = 3radic31199031198902 (1 minus cos (120579)) (4)

rc




rceil (5)









Num

ber o

f sta

tions

100

101

102

103

104

105

106

107
























Sink 1

Thing 1Thing 1

Things

Things

Router 1

Sink 1

Thing 1Thing 1

ThingsSink n

Thing 1Thing 1

Router n

Internet things


Thing 1

Thing 2 Thing 3

Thing n

HAP 1 HAP n

Thing 1

Things







Sink 1

Thing 1Thing 1

ThingsThings

Router 1

Sink 1

Thing 1Thing 1

ThingsSink n

Thing 1Thing 1

Router n

Internet things

Thing 1

Thing 2 Thing 3

Thing n

HAP 1 HAP n

Thing 1

Things


25milliseconds15Mbs


Sink 1

Thing 1Thing 1


Sink 1

Thing 1Thing 1

Sink n

Thing 1Thing 1

Router 1 Router n

HAP 1 HAP n


things

Random speed

Random speed

Random speed

WLAN link

WLAN link

WLAN link

25milliseconds15Mbs

25milliseconds15Mbs


2Mbs 2Mbs1Mbs








0

10

20

30

40

50

60

70

80

90

100

1 10 19 28 37 46 55 64 73 82 91 100

Aver

age d

elay

(ms)




0

20

40

60

80

100

120

1 10 19 28 37 46 55 64 73 82 91 100

Aver

age d

elay

(ms)







0

002

004

006

008

01

012

014

016

018

02

1 10 19 28 37 46 55 64 73 82 91 100

Pack

et lo

ss ra

tio




0

01

02

03

04

05

06

07

1 10 19 28 37 46 55 64 73 82 91 100

Pack

et lo

ss ra

tio








0100000020000003000000400000050000006000000700000080000009000000

1000000011000000

1 10 19 28 37 46 55 64 73 82 91 100

Thro

ughp

ut (b

itss

ec) (

10)






0

1000000

2000000

3000000

4000000

5000000

6000000

7000000

8000000

1 10 19 28 37 46 55 64 73 82 91 100

Thro

ughp

ut (b

itss

ec) (

10)








0

100

200

300

400

500

600

Aver

age o

f ene

rgy

cons

umpt

ion

(J)




0

50

100

150

200

250

300

350

400

Aver

age o

f ene

rgy

cons

umpt

ion

(J)




0

20

40

60

80

100

120

140

Aver

age o

f ene

rgy

cons

umpt

ion

(J)









0

100

200

300

400

500

600

700

Aver

age o

f ene

rgy

cons

umpt

ion

(J)




050

100150200250300350400450500

Aver

age o

f ene

rgy

cons

umpt

ion

(J)




020406080

100120140160180

Aver

age o

f ene

rgy

cons

umpt

ion

(J)








0

01

02

03

04

05

06

07

08

09

0 10 20 30 40 50 60 70 80 90 100 110 120 130 140 150

Han

dove

r pro

babi

lity

Speed (kmh)



0

002

004

006

008

01

012

014

016

018

02

022

024

0 10 20 30 40 50 60 70 80 90 100 110 120 130 140 150

Bloc

king

pro

babi

lity

Speed (kmh)




0

005

01

015

02

025

03

035

0 10 20 30 40 50 60 70 80 90 100 110 120 130 140 150

Dro

ppin

g pr

obab

ility

(11

000)

Speed (kmh)



7 Conclusion


8 Future Work




Competing Interests


Acknowledgments


References









































































Advances in

FuzzySystems


Volume 2014












Journal of

Journal of





Advances in

Multimedia


Biomedical Imaging



Advances in


RoboticsJournal of










Advances in




First layerHAP with

space things




Backhaul link


First layer

HAP

Second layer

HAPs




Backhaul link



119860 = 21205871199031198902 (1 minus cos (120579)) (1)





First layerHAP with

space things





Backhaul link


SatelliteHAP

h

E

e

re



119886ℎ = 3radic32 1199031198882 (3)


119886ℎ = 3radic31199031198902 (1 minus cos (120579)) (4)

rc




rceil (5)









Num

ber o

f sta

tions

100

101

102

103

104

105

106

107
























Sink 1

Thing 1Thing 1

Things

Things

Router 1

Sink 1

Thing 1Thing 1

ThingsSink n

Thing 1Thing 1

Router n

Internet things


Thing 1

Thing 2 Thing 3

Thing n

HAP 1 HAP n

Thing 1

Things







Sink 1

Thing 1Thing 1

ThingsThings

Router 1

Sink 1

Thing 1Thing 1

ThingsSink n

Thing 1Thing 1

Router n

Internet things

Thing 1

Thing 2 Thing 3

Thing n

HAP 1 HAP n

Thing 1

Things


25milliseconds15Mbs


Sink 1

Thing 1Thing 1


Sink 1

Thing 1Thing 1

Sink n

Thing 1Thing 1

Router 1 Router n

HAP 1 HAP n


things

Random speed

Random speed

Random speed

WLAN link

WLAN link

WLAN link

25milliseconds15Mbs

25milliseconds15Mbs


2Mbs 2Mbs1Mbs








0

10

20

30

40

50

60

70

80

90

100

1 10 19 28 37 46 55 64 73 82 91 100

Aver

age d

elay

(ms)




0

20

40

60

80

100

120

1 10 19 28 37 46 55 64 73 82 91 100

Aver

age d

elay

(ms)







0

002

004

006

008

01

012

014

016

018

02

1 10 19 28 37 46 55 64 73 82 91 100

Pack

et lo

ss ra

tio




0

01

02

03

04

05

06

07

1 10 19 28 37 46 55 64 73 82 91 100

Pack

et lo

ss ra

tio








0100000020000003000000400000050000006000000700000080000009000000

1000000011000000

1 10 19 28 37 46 55 64 73 82 91 100

Thro

ughp

ut (b

itss

ec) (

10)






0

1000000

2000000

3000000

4000000

5000000

6000000

7000000

8000000

1 10 19 28 37 46 55 64 73 82 91 100

Thro

ughp

ut (b

itss

ec) (

10)








0

100

200

300

400

500

600

Aver

age o

f ene

rgy

cons

umpt

ion

(J)




0

50

100

150

200

250

300

350

400

Aver

age o

f ene

rgy

cons

umpt

ion

(J)




0

20

40

60

80

100

120

140

Aver

age o

f ene

rgy

cons

umpt

ion

(J)









0

100

200

300

400

500

600

700

Aver

age o

f ene

rgy

cons

umpt

ion

(J)




050

100150200250300350400450500

Aver

age o

f ene

rgy

cons

umpt

ion

(J)




020406080

100120140160180

Aver

age o

f ene

rgy

cons

umpt

ion

(J)








0

01

02

03

04

05

06

07

08

09

0 10 20 30 40 50 60 70 80 90 100 110 120 130 140 150

Han

dove

r pro

babi

lity

Speed (kmh)



0

002

004

006

008

01

012

014

016

018

02

022

024

0 10 20 30 40 50 60 70 80 90 100 110 120 130 140 150

Bloc

king

pro

babi

lity

Speed (kmh)




0

005

01

015

02

025

03

035

0 10 20 30 40 50 60 70 80 90 100 110 120 130 140 150

Dro

ppin

g pr

obab

ility

(11

000)

Speed (kmh)



7 Conclusion


8 Future Work




Competing Interests


Acknowledgments


References









































































Advances in

FuzzySystems


Volume 2014












Journal of

Journal of





Advances in

Multimedia


Biomedical Imaging



Advances in


RoboticsJournal of










Advances in




First layerHAP with

space things





Backhaul link


SatelliteHAP

h

E

e

re



119886ℎ = 3radic32 1199031198882 (3)


119886ℎ = 3radic31199031198902 (1 minus cos (120579)) (4)

rc




rceil (5)









Num

ber o

f sta

tions

100

101

102

103

104

105

106

107
























Sink 1

Thing 1Thing 1

Things

Things

Router 1

Sink 1

Thing 1Thing 1

ThingsSink n

Thing 1Thing 1

Router n

Internet things


Thing 1

Thing 2 Thing 3

Thing n

HAP 1 HAP n

Thing 1

Things







Sink 1

Thing 1Thing 1

ThingsThings

Router 1

Sink 1

Thing 1Thing 1

ThingsSink n

Thing 1Thing 1

Router n

Internet things

Thing 1

Thing 2 Thing 3

Thing n

HAP 1 HAP n

Thing 1

Things


25milliseconds15Mbs


Sink 1

Thing 1Thing 1


Sink 1

Thing 1Thing 1

Sink n

Thing 1Thing 1

Router 1 Router n

HAP 1 HAP n


things

Random speed

Random speed

Random speed

WLAN link

WLAN link

WLAN link

25milliseconds15Mbs

25milliseconds15Mbs


2Mbs 2Mbs1Mbs








0

10

20

30

40

50

60

70

80

90

100

1 10 19 28 37 46 55 64 73 82 91 100

Aver

age d

elay

(ms)




0

20

40

60

80

100

120

1 10 19 28 37 46 55 64 73 82 91 100

Aver

age d

elay

(ms)







0

002

004

006

008

01

012

014

016

018

02

1 10 19 28 37 46 55 64 73 82 91 100

Pack

et lo

ss ra

tio




0

01

02

03

04

05

06

07

1 10 19 28 37 46 55 64 73 82 91 100

Pack

et lo

ss ra

tio








0100000020000003000000400000050000006000000700000080000009000000

1000000011000000

1 10 19 28 37 46 55 64 73 82 91 100

Thro

ughp

ut (b

itss

ec) (

10)






0

1000000

2000000

3000000

4000000

5000000

6000000

7000000

8000000

1 10 19 28 37 46 55 64 73 82 91 100

Thro

ughp

ut (b

itss

ec) (

10)








0

100

200

300

400

500

600

Aver

age o

f ene

rgy

cons

umpt

ion

(J)




0

50

100

150

200

250

300

350

400

Aver

age o

f ene

rgy

cons

umpt

ion

(J)




0

20

40

60

80

100

120

140

Aver

age o

f ene

rgy

cons

umpt

ion

(J)









0

100

200

300

400

500

600

700

Aver

age o

f ene

rgy

cons

umpt

ion

(J)




050

100150200250300350400450500

Aver

age o

f ene

rgy

cons

umpt

ion

(J)




020406080

100120140160180

Aver

age o

f ene

rgy

cons

umpt

ion

(J)








0

01

02

03

04

05

06

07

08

09

0 10 20 30 40 50 60 70 80 90 100 110 120 130 140 150

Han

dove

r pro

babi

lity

Speed (kmh)



0

002

004

006

008

01

012

014

016

018

02

022

024

0 10 20 30 40 50 60 70 80 90 100 110 120 130 140 150

Bloc

king

pro

babi

lity

Speed (kmh)




0

005

01

015

02

025

03

035

0 10 20 30 40 50 60 70 80 90 100 110 120 130 140 150

Dro

ppin

g pr

obab

ility

(11

000)

Speed (kmh)



7 Conclusion


8 Future Work




Competing Interests


Acknowledgments


References









































































Advances in

FuzzySystems


Volume 2014












Journal of

Journal of





Advances in

Multimedia


Biomedical Imaging



Advances in


RoboticsJournal of










Advances in









Num

ber o

f sta

tions

100

101

102

103

104

105

106

107
























Sink 1

Thing 1Thing 1

Things

Things

Router 1

Sink 1

Thing 1Thing 1

ThingsSink n

Thing 1Thing 1

Router n

Internet things


Thing 1

Thing 2 Thing 3

Thing n

HAP 1 HAP n

Thing 1

Things







Sink 1

Thing 1Thing 1

ThingsThings

Router 1

Sink 1

Thing 1Thing 1

ThingsSink n

Thing 1Thing 1

Router n

Internet things

Thing 1

Thing 2 Thing 3

Thing n

HAP 1 HAP n

Thing 1

Things


25milliseconds15Mbs


Sink 1

Thing 1Thing 1


Sink 1

Thing 1Thing 1

Sink n

Thing 1Thing 1

Router 1 Router n

HAP 1 HAP n


things

Random speed

Random speed

Random speed

WLAN link

WLAN link

WLAN link

25milliseconds15Mbs

25milliseconds15Mbs


2Mbs 2Mbs1Mbs








0

10

20

30

40

50

60

70

80

90

100

1 10 19 28 37 46 55 64 73 82 91 100

Aver

age d

elay

(ms)




0

20

40

60

80

100

120

1 10 19 28 37 46 55 64 73 82 91 100

Aver

age d

elay

(ms)







0

002

004

006

008

01

012

014

016

018

02

1 10 19 28 37 46 55 64 73 82 91 100

Pack

et lo

ss ra

tio




0

01

02

03

04

05

06

07

1 10 19 28 37 46 55 64 73 82 91 100

Pack

et lo

ss ra

tio








0100000020000003000000400000050000006000000700000080000009000000

1000000011000000

1 10 19 28 37 46 55 64 73 82 91 100

Thro

ughp

ut (b

itss

ec) (

10)






0

1000000

2000000

3000000

4000000

5000000

6000000

7000000

8000000

1 10 19 28 37 46 55 64 73 82 91 100

Thro

ughp

ut (b

itss

ec) (

10)








0

100

200

300

400

500

600

Aver

age o

f ene

rgy

cons

umpt

ion

(J)




0

50

100

150

200

250

300

350

400

Aver

age o

f ene

rgy

cons

umpt

ion

(J)




0

20

40

60

80

100

120

140

Aver

age o

f ene

rgy

cons

umpt

ion

(J)









0

100

200

300

400

500

600

700

Aver

age o

f ene

rgy

cons

umpt

ion

(J)




050

100150200250300350400450500

Aver

age o

f ene

rgy

cons

umpt

ion

(J)




020406080

100120140160180

Aver

age o

f ene

rgy

cons

umpt

ion

(J)








0

01

02

03

04

05

06

07

08

09

0 10 20 30 40 50 60 70 80 90 100 110 120 130 140 150

Han

dove

r pro

babi

lity

Speed (kmh)



0

002

004

006

008

01

012

014

016

018

02

022

024

0 10 20 30 40 50 60 70 80 90 100 110 120 130 140 150

Bloc

king

pro

babi

lity

Speed (kmh)




0

005

01

015

02

025

03

035

0 10 20 30 40 50 60 70 80 90 100 110 120 130 140 150

Dro

ppin

g pr

obab

ility

(11

000)

Speed (kmh)



7 Conclusion


8 Future Work




Competing Interests


Acknowledgments


References









































































Advances in

FuzzySystems


Volume 2014












Journal of

Journal of





Advances in

Multimedia


Biomedical Imaging



Advances in


RoboticsJournal of










Advances in


















Sink 1

Thing 1Thing 1

Things

Things

Router 1

Sink 1

Thing 1Thing 1

ThingsSink n

Thing 1Thing 1

Router n

Internet things


Thing 1

Thing 2 Thing 3

Thing n

HAP 1 HAP n

Thing 1

Things







Sink 1

Thing 1Thing 1

ThingsThings

Router 1

Sink 1

Thing 1Thing 1

ThingsSink n

Thing 1Thing 1

Router n

Internet things

Thing 1

Thing 2 Thing 3

Thing n

HAP 1 HAP n

Thing 1

Things


25milliseconds15Mbs


Sink 1

Thing 1Thing 1


Sink 1

Thing 1Thing 1

Sink n

Thing 1Thing 1

Router 1 Router n

HAP 1 HAP n


things

Random speed

Random speed

Random speed

WLAN link

WLAN link

WLAN link

25milliseconds15Mbs

25milliseconds15Mbs


2Mbs 2Mbs1Mbs








0

10

20

30

40

50

60

70

80

90

100

1 10 19 28 37 46 55 64 73 82 91 100

Aver

age d

elay

(ms)




0

20

40

60

80

100

120

1 10 19 28 37 46 55 64 73 82 91 100

Aver

age d

elay

(ms)







0

002

004

006

008

01

012

014

016

018

02

1 10 19 28 37 46 55 64 73 82 91 100

Pack

et lo

ss ra

tio




0

01

02

03

04

05

06

07

1 10 19 28 37 46 55 64 73 82 91 100

Pack

et lo

ss ra

tio








0100000020000003000000400000050000006000000700000080000009000000

1000000011000000

1 10 19 28 37 46 55 64 73 82 91 100

Thro

ughp

ut (b

itss

ec) (

10)






0

1000000

2000000

3000000

4000000

5000000

6000000

7000000

8000000

1 10 19 28 37 46 55 64 73 82 91 100

Thro

ughp

ut (b

itss

ec) (

10)








0

100

200

300

400

500

600

Aver

age o

f ene

rgy

cons

umpt

ion

(J)




0

50

100

150

200

250

300

350

400

Aver

age o

f ene

rgy

cons

umpt

ion

(J)




0

20

40

60

80

100

120

140

Aver

age o

f ene

rgy

cons

umpt

ion

(J)









0

100

200

300

400

500

600

700

Aver

age o

f ene

rgy

cons

umpt

ion

(J)




050

100150200250300350400450500

Aver

age o

f ene

rgy

cons

umpt

ion

(J)




020406080

100120140160180

Aver

age o

f ene

rgy

cons

umpt

ion

(J)








0

01

02

03

04

05

06

07

08

09

0 10 20 30 40 50 60 70 80 90 100 110 120 130 140 150

Han

dove

r pro

babi

lity

Speed (kmh)



0

002

004

006

008

01

012

014

016

018

02

022

024

0 10 20 30 40 50 60 70 80 90 100 110 120 130 140 150

Bloc

king

pro

babi

lity

Speed (kmh)




0

005

01

015

02

025

03

035

0 10 20 30 40 50 60 70 80 90 100 110 120 130 140 150

Dro

ppin

g pr

obab

ility

(11

000)

Speed (kmh)



7 Conclusion


8 Future Work




Competing Interests


Acknowledgments


References









































































Advances in

FuzzySystems


Volume 2014












Journal of

Journal of





Advances in

Multimedia


Biomedical Imaging



Advances in


RoboticsJournal of










Advances in





Sink 1

Thing 1Thing 1

ThingsThings

Router 1

Sink 1

Thing 1Thing 1

ThingsSink n

Thing 1Thing 1

Router n

Internet things

Thing 1

Thing 2 Thing 3

Thing n

HAP 1 HAP n

Thing 1

Things


25milliseconds15Mbs


Sink 1

Thing 1Thing 1


Sink 1

Thing 1Thing 1

Sink n

Thing 1Thing 1

Router 1 Router n

HAP 1 HAP n


things

Random speed

Random speed

Random speed

WLAN link

WLAN link

WLAN link

25milliseconds15Mbs

25milliseconds15Mbs


2Mbs 2Mbs1Mbs








0

10

20

30

40

50

60

70

80

90

100

1 10 19 28 37 46 55 64 73 82 91 100

Aver

age d

elay

(ms)




0

20

40

60

80

100

120

1 10 19 28 37 46 55 64 73 82 91 100

Aver

age d

elay

(ms)







0

002

004

006

008

01

012

014

016

018

02

1 10 19 28 37 46 55 64 73 82 91 100

Pack

et lo

ss ra

tio




0

01

02

03

04

05

06

07

1 10 19 28 37 46 55 64 73 82 91 100

Pack

et lo

ss ra

tio








0100000020000003000000400000050000006000000700000080000009000000

1000000011000000

1 10 19 28 37 46 55 64 73 82 91 100

Thro

ughp

ut (b

itss

ec) (

10)






0

1000000

2000000

3000000

4000000

5000000

6000000

7000000

8000000

1 10 19 28 37 46 55 64 73 82 91 100

Thro

ughp

ut (b

itss

ec) (

10)








0

100

200

300

400

500

600

Aver

age o

f ene

rgy

cons

umpt

ion

(J)




0

50

100

150

200

250

300

350

400

Aver

age o

f ene

rgy

cons

umpt

ion

(J)




0

20

40

60

80

100

120

140

Aver

age o

f ene

rgy

cons

umpt

ion

(J)









0

100

200

300

400

500

600

700

Aver

age o

f ene

rgy

cons

umpt

ion

(J)




050

100150200250300350400450500

Aver

age o

f ene

rgy

cons

umpt

ion

(J)




020406080

100120140160180

Aver

age o

f ene

rgy

cons

umpt

ion

(J)








0

01

02

03

04

05

06

07

08

09

0 10 20 30 40 50 60 70 80 90 100 110 120 130 140 150

Han

dove

r pro

babi

lity

Speed (kmh)



0

002

004

006

008

01

012

014

016

018

02

022

024

0 10 20 30 40 50 60 70 80 90 100 110 120 130 140 150

Bloc

king

pro

babi

lity

Speed (kmh)




0

005

01

015

02

025

03

035

0 10 20 30 40 50 60 70 80 90 100 110 120 130 140 150

Dro

ppin

g pr

obab

ility

(11

000)

Speed (kmh)



7 Conclusion


8 Future Work




Competing Interests


Acknowledgments


References









































































Advances in

FuzzySystems


Volume 2014












Journal of

Journal of





Advances in

Multimedia


Biomedical Imaging



Advances in


RoboticsJournal of










Advances in







0

10

20

30

40

50

60

70

80

90

100

1 10 19 28 37 46 55 64 73 82 91 100

Aver

age d

elay

(ms)




0

20

40

60

80

100

120

1 10 19 28 37 46 55 64 73 82 91 100

Aver

age d

elay

(ms)







0

002

004

006

008

01

012

014

016

018

02

1 10 19 28 37 46 55 64 73 82 91 100

Pack

et lo

ss ra

tio




0

01

02

03

04

05

06

07

1 10 19 28 37 46 55 64 73 82 91 100

Pack

et lo

ss ra

tio








0100000020000003000000400000050000006000000700000080000009000000

1000000011000000

1 10 19 28 37 46 55 64 73 82 91 100

Thro

ughp

ut (b

itss

ec) (

10)






0

1000000

2000000

3000000

4000000

5000000

6000000

7000000

8000000

1 10 19 28 37 46 55 64 73 82 91 100

Thro

ughp

ut (b

itss

ec) (

10)








0

100

200

300

400

500

600

Aver

age o

f ene

rgy

cons

umpt

ion

(J)




0

50

100

150

200

250

300

350

400

Aver

age o

f ene

rgy

cons

umpt

ion

(J)




0

20

40

60

80

100

120

140

Aver

age o

f ene

rgy

cons

umpt

ion

(J)









0

100

200

300

400

500

600

700

Aver

age o

f ene

rgy

cons

umpt

ion

(J)




050

100150200250300350400450500

Aver

age o

f ene

rgy

cons

umpt

ion

(J)




020406080

100120140160180

Aver

age o

f ene

rgy

cons

umpt

ion

(J)








0

01

02

03

04

05

06

07

08

09

0 10 20 30 40 50 60 70 80 90 100 110 120 130 140 150

Han

dove

r pro

babi

lity

Speed (kmh)



0

002

004

006

008

01

012

014

016

018

02

022

024

0 10 20 30 40 50 60 70 80 90 100 110 120 130 140 150

Bloc

king

pro

babi

lity

Speed (kmh)




0

005

01

015

02

025

03

035

0 10 20 30 40 50 60 70 80 90 100 110 120 130 140 150

Dro

ppin

g pr

obab

ility

(11

000)

Speed (kmh)



7 Conclusion


8 Future Work




Competing Interests


Acknowledgments


References









































































Advances in

FuzzySystems


Volume 2014












Journal of

Journal of





Advances in

Multimedia


Biomedical Imaging



Advances in


RoboticsJournal of










Advances in




0

002

004

006

008

01

012

014

016

018

02

1 10 19 28 37 46 55 64 73 82 91 100

Pack

et lo

ss ra

tio




0

01

02

03

04

05

06

07

1 10 19 28 37 46 55 64 73 82 91 100

Pack

et lo

ss ra

tio








0100000020000003000000400000050000006000000700000080000009000000

1000000011000000

1 10 19 28 37 46 55 64 73 82 91 100

Thro

ughp

ut (b

itss

ec) (

10)






0

1000000

2000000

3000000

4000000

5000000

6000000

7000000

8000000

1 10 19 28 37 46 55 64 73 82 91 100

Thro

ughp

ut (b

itss

ec) (

10)








0

100

200

300

400

500

600

Aver

age o

f ene

rgy

cons

umpt

ion

(J)




0

50

100

150

200

250

300

350

400

Aver

age o

f ene

rgy

cons

umpt

ion

(J)




0

20

40

60

80

100

120

140

Aver

age o

f ene

rgy

cons

umpt

ion

(J)









0

100

200

300

400

500

600

700

Aver

age o

f ene

rgy

cons

umpt

ion

(J)




050

100150200250300350400450500

Aver

age o

f ene

rgy

cons

umpt

ion

(J)




020406080

100120140160180

Aver

age o

f ene

rgy

cons

umpt

ion

(J)








0

01

02

03

04

05

06

07

08

09

0 10 20 30 40 50 60 70 80 90 100 110 120 130 140 150

Han

dove

r pro

babi

lity

Speed (kmh)



0

002

004

006

008

01

012

014

016

018

02

022

024

0 10 20 30 40 50 60 70 80 90 100 110 120 130 140 150

Bloc

king

pro

babi

lity

Speed (kmh)




0

005

01

015

02

025

03

035

0 10 20 30 40 50 60 70 80 90 100 110 120 130 140 150

Dro

ppin

g pr

obab

ility

(11

000)

Speed (kmh)



7 Conclusion


8 Future Work




Competing Interests


Acknowledgments


References









































































Advances in

FuzzySystems


Volume 2014












Journal of

Journal of





Advances in

Multimedia


Biomedical Imaging



Advances in


RoboticsJournal of










Advances in




0100000020000003000000400000050000006000000700000080000009000000

1000000011000000

1 10 19 28 37 46 55 64 73 82 91 100

Thro

ughp

ut (b

itss

ec) (

10)






0

1000000

2000000

3000000

4000000

5000000

6000000

7000000

8000000

1 10 19 28 37 46 55 64 73 82 91 100

Thro

ughp

ut (b

itss

ec) (

10)








0

100

200

300

400

500

600

Aver

age o

f ene

rgy

cons

umpt

ion

(J)




0

50

100

150

200

250

300

350

400

Aver

age o

f ene

rgy

cons

umpt

ion

(J)




0

20

40

60

80

100

120

140

Aver

age o

f ene

rgy

cons

umpt

ion

(J)









0

100

200

300

400

500

600

700

Aver

age o

f ene

rgy

cons

umpt

ion

(J)




050

100150200250300350400450500

Aver

age o

f ene

rgy

cons

umpt

ion

(J)




020406080

100120140160180

Aver

age o

f ene

rgy

cons

umpt

ion

(J)








0

01

02

03

04

05

06

07

08

09

0 10 20 30 40 50 60 70 80 90 100 110 120 130 140 150

Han

dove

r pro

babi

lity

Speed (kmh)



0

002

004

006

008

01

012

014

016

018

02

022

024

0 10 20 30 40 50 60 70 80 90 100 110 120 130 140 150

Bloc

king

pro

babi

lity

Speed (kmh)




0

005

01

015

02

025

03

035

0 10 20 30 40 50 60 70 80 90 100 110 120 130 140 150

Dro

ppin

g pr

obab

ility

(11

000)

Speed (kmh)



7 Conclusion


8 Future Work




Competing Interests


Acknowledgments


References









































































Advances in

FuzzySystems


Volume 2014












Journal of

Journal of





Advances in

Multimedia


Biomedical Imaging



Advances in


RoboticsJournal of










Advances in





0

100

200

300

400

500

600

Aver

age o

f ene

rgy

cons

umpt

ion

(J)




0

50

100

150

200

250

300

350

400

Aver

age o

f ene

rgy

cons

umpt

ion

(J)




0

20

40

60

80

100

120

140

Aver

age o

f ene

rgy

cons

umpt

ion

(J)









0

100

200

300

400

500

600

700

Aver

age o

f ene

rgy

cons

umpt

ion

(J)




050

100150200250300350400450500

Aver

age o

f ene

rgy

cons

umpt

ion

(J)




020406080

100120140160180

Aver

age o

f ene

rgy

cons

umpt

ion

(J)








0

01

02

03

04

05

06

07

08

09

0 10 20 30 40 50 60 70 80 90 100 110 120 130 140 150

Han

dove

r pro

babi

lity

Speed (kmh)



0

002

004

006

008

01

012

014

016

018

02

022

024

0 10 20 30 40 50 60 70 80 90 100 110 120 130 140 150

Bloc

king

pro

babi

lity

Speed (kmh)




0

005

01

015

02

025

03

035

0 10 20 30 40 50 60 70 80 90 100 110 120 130 140 150

Dro

ppin

g pr

obab

ility

(11

000)

Speed (kmh)



7 Conclusion


8 Future Work




Competing Interests


Acknowledgments


References









































































Advances in

FuzzySystems


Volume 2014












Journal of

Journal of





Advances in

Multimedia


Biomedical Imaging



Advances in


RoboticsJournal of










Advances in





0

100

200

300

400

500

600

700

Aver

age o

f ene

rgy

cons

umpt

ion

(J)




050

100150200250300350400450500

Aver

age o

f ene

rgy

cons

umpt

ion

(J)




020406080

100120140160180

Aver

age o

f ene

rgy

cons

umpt

ion

(J)








0

01

02

03

04

05

06

07

08

09

0 10 20 30 40 50 60 70 80 90 100 110 120 130 140 150

Han

dove

r pro

babi

lity

Speed (kmh)



0

002

004

006

008

01

012

014

016

018

02

022

024

0 10 20 30 40 50 60 70 80 90 100 110 120 130 140 150

Bloc

king

pro

babi

lity

Speed (kmh)




0

005

01

015

02

025

03

035

0 10 20 30 40 50 60 70 80 90 100 110 120 130 140 150

Dro

ppin

g pr

obab

ility

(11

000)

Speed (kmh)



7 Conclusion


8 Future Work




Competing Interests


Acknowledgments


References









































































Advances in

FuzzySystems


Volume 2014












Journal of

Journal of





Advances in

Multimedia


Biomedical Imaging



Advances in


RoboticsJournal of










Advances in




0

01

02

03

04

05

06

07

08

09

0 10 20 30 40 50 60 70 80 90 100 110 120 130 140 150

Han

dove

r pro

babi

lity

Speed (kmh)



0

002

004

006

008

01

012

014

016

018

02

022

024

0 10 20 30 40 50 60 70 80 90 100 110 120 130 140 150

Bloc

king

pro

babi

lity

Speed (kmh)




0

005

01

015

02

025

03

035

0 10 20 30 40 50 60 70 80 90 100 110 120 130 140 150

Dro

ppin

g pr

obab

ility

(11

000)

Speed (kmh)



7 Conclusion


8 Future Work




Competing Interests


Acknowledgments


References









































































Advances in

FuzzySystems


Volume 2014












Journal of

Journal of





Advances in

Multimedia


Biomedical Imaging



Advances in


RoboticsJournal of










Advances in





Competing Interests


Acknowledgments


References









































































Advances in

FuzzySystems


Volume 2014












Journal of

Journal of





Advances in

Multimedia


Biomedical Imaging



Advances in


RoboticsJournal of










Advances in



















































Advances in

FuzzySystems


Volume 2014












Journal of

Journal of





Advances in

Multimedia


Biomedical Imaging



Advances in


RoboticsJournal of










Advances in



research article performance evaluation of a dual...

Documents