threshold balanced sampled deec model for heterogeneous...

Research ArticleThreshold Balanced Sampled DEEC Model for HeterogeneousWireless Sensor Networks

Sercan Vanccedilin and Ebubekir Erdem

Department of Computer Engineering Firat University Elazig 23100 Turkey

Correspondence should be addressed to Ebubekir Erdem aberdemfiratedutr

Received 23 March 2018 Revised 31 August 2018 Accepted 10 October 2018 Published 6 November 2018

Academic Editor Luca De Nardis

Copyright copy 2018 Sercan Vancin and Ebubekir Erdem This is an open access article distributed under the Creative CommonsAttribution License which permits unrestricted use distribution and reproduction in any medium provided the original work isproperly cited

Due to the restricted hardware resources of the sensor nodes modelling and designing energy efficient routing methods to increasethe overall network lifetime have become one of the most significant strategies in wireless sensor networks (WSNs) Cluster-basedheterogeneous routing protocols a popular part of routing technology have proven effective in management of topology energyconsumption data collection or fusion reliability or stability in a distributed sensor network In this article an energy efficientthree-level heterogeneous clustering method (DEEC) based distributed energy efficient clustering protocol named TBSDEEC(Threshold balanced sampled DEEC) is proposed Contrary to most other studies this study considers the effect of the thresholdbalanced sampled in the energy consumption model Our model is compared with the DEEC EDEEC (Enhanced DistributedEnergy Efficient Clustering Protocol) and EDDEEC (EnhancedDevelopedDistributed Energy Efficient Clustering Protocol) usingMATLAB as two different scenarios based on qualitymetrics including living nodes on the network network efficiency energy con-sumption number of packets received by base station (BS) and average latency After our new method is compared with artificialbee colony optimization (ABCO) algorithm and energy harvestingWSN (EH-WSN) clusteringmethod Simulation results demon-strate that the proposed model is more efficient than the other protocols and significantly increases the sensor network lifetime

1 Introduction

Wireless sensor networks (WSNs) include small-sized sensornodes that can transmit data through data sensing compu-tation and wireless channel communication capabilities [1]One of the major problems in the WSN is the limited batterypower at the sensor nodes Routing protocols around thework areas of the WSN are an important area In additionto prolonging the life of the sensor nodes it is also desirableto distribute the existing energy homogeneously to theWSNDue to the limited power supply in the sensor nodes theenergy consumption of the power source is an importantconcept in the WSNs Maximum energy is used when datais transmitted to other nodes via sensor nodes For all thesereasons a number of studies have been conducted to developrouting algorithms to extend a sensor network lifetime [2]

11 Problem Definition To extend the lifespan of sensornetworks sensor networks are on the basis of the common-ality of sensors by providing energy saving and scalability

called aggregation In this sense nodes in the WSN work incooperation by separating into clusters [3] However in a fewliterature studies the advantages of distributing energy in amore balanced way over the network have been discussedBy the heterogeneous structure of the nodes in the networkwhether the nodes are close to BS or not different clustersize structures are the issues that should be taken as basis Insome studies only homogeneous nodes were used while insome studies the concept of distributed energy was not takeninto account Our aim in this study is to consider all theseproblems in an integratedway to overcome themand tomakeenergy efficient networks more qualified

12 RelatedWorks In the literature there aremany studies onenergy efficient clustering of protocols forWSNs In one study[4] a routing algorithm with LEACH clustering adaptation ispresented for homogeneous WSNs where sensor nodes arerandomly determined as CHs (Cluster Heads) and the energyload of the system is shared with the WSN In [5] a new

HindawiWireless Communications and Mobile ComputingVolume 2018 Article ID 4618056 12 pageshttpsdoiorg10115520184618056

2 Wireless Communications and Mobile Computing

routing protocol based on LEACH for energy optimizationis proposed It is understood that this algorithm is moreefficient than the LEACH algorithmby selecting cluster headsequally The paper [6] presents a modified LEACH derivedfrom the LEACH algorithm In [7] a mobile sink improvedenergy efficient algorithm is presented and compared withmod-LEACH and PEGASIS [8] In [9] a new energy efficient(EE) clustering based method is proposed for single passheterogeneousWSNs Simulations inMATLAB show that thementioned method has a 162-189 times better stability thanknown protocols such as LEACHDEEC and SEP In [10] thestability of the cluster is reduced because the LEACHprotocolon an irregular network causes a decrease in aggregate dataefficiency For this reason this article [10] suggests a methodof selecting a cluster head to improve the LEACH protocol inorder to increase cluster head stability For this purpose anLEACH variant combined with HEED and LEACH protocolis proposed and this method is approved by simulation In[11] two energy efficient route planning routing protocols areproposed for three levels of heterogeneous WSNs namelyCentral Energy Efficiency Clustering (CEEC) with Two-Hop Heterogeneity awareness (THCEEC) and AdvancedEqualization (ACEEC) Comprehensive simulation resultshave provided CEEC ACEEC and THCEEC central clusterdeployments with improved reliability and energy efficiencyperformance providing better network lifetime and success-ful data transmission than LEEC SEP ESEP and DEECrsquostraditional distributed routing protocols In addition ACEECperforms CEEC and provides more network stability timeAnalytical evaluation shows that THCEEC performs CEECACEEC and other existing road planning routing protocolsThe study [12] suggests an efficient method of collecting datawith a support vector in the WSN In [13] performanceevaluation of clustering protocols is presented in WSNsClustering of sensor nodes is an effective technique inreaching these targets With this technique other clusteringmodels (LEACH LEACH-C and HEED) were evaluated andcompared At the end of these clustering methods are com-pared with depending on several criteria such as convergencespeed cluster stability cluster overlap location awarenessand node mobility support In another study [14] the studyof various routing models for sensor networks offers a surveywith classification on behalf of kinds of models The threemain categories examined are data-centric hierarchical andlocation-based Routing methods and algorithms each havea common purpose to better output and extend the usefullife of the sensor network A comparison was made betweenflood and direct diffusion two routing protocols based onnetwork throughput and lifetime Simulation of AODV (AdHoc onDemand Distance Vector) was also performed on twotopologies with the same source and target nodes The study[15] presents random coverage and connectivity analysis inheterogeneous WSNs with three-dimensions The study [16]suggests that the SEP algorithm in which each sensor nodein a two-level heterogeneous sensor network independentlyidentifies itself as a CH on the basis of the first energyrelative to the other sensor nodes of the sensor networkThe study [17] presents a method named DEEC by whichthe CH selection is considered to depend on the ratio of

the remaining energy of the node and the average energyof the sensor network In a study [18] the DDEEC protocolis presented based on the recalibration of the energy forCH This protocol has been optimized by the public wirelessnetwork In this sense it is more likely that advanced nodeswill be chosen as CH in the first broadcast rounds Also whenenergy is reduced these sensor nodes will have the sameprobability of CH selection as normal sensor nodesThe study[19] demonstrates EDEEC a clustering method with a three-level heterogeneous structure that yields a high amount ofenergy level called super sensor nodes In one study [20] aclustering protocol known as EDDEEC was introduced CHselection probability is dependent on the remaining energyquality of the sensor nodes with the average energy of theWSN In one study the opportunity of each node to be chosenas a CH is determined towards to its energy level and to theamount of depleted energy Nodes with higher opportunityhave less delay times The node with the smallest time delaycomparing to its neighbours is chosen as CH After choosinga CH and forming a cluster all nodes of each cluster beginto send packets to the CH depend on energy-aware multihoprouting Then CH sends these packets to the BS by means ofmultihop routing [21] The aim of paper [22] is to analyse theperformance of artificial bee colony optimization algorithm(ABCO) depending on clusteringmethod utilized to improvethe network lifetime The node with highest energy in acluster is chosen as CH in a determined period and the allfield is reclustered depending on the selected CHs In onepaper [23] an energy efficient routing protocol for wirelesssensor networks is recommended This method consists of arouting algorithm for the transmission of data CH choosingalgorithm and a scheme for the formation of clusters

13 Contributions and Motivations Our main contributionscan be listed as follows

(i) The proposed method (TBSDEEC) is a DEEC-basedenergy efficient three-level heterogeneous clusteringmodel named TBSDEEC for distributed sensor net-works

(ii) The proposed method uses an EDDEEC-like net-workmodel However while the energy consumptionmodel is designing the threshold energy level differsfrom other protocols

(iii) In this study we describe a balanced value calledthreshold balanced sample energy (TBSE) which wasproposed and contributed to the calculation of thethreshold value

(iv) Contrary to other studies the model we proposedprovides more accurate and precise solutions for theselection of the CH and the threshold value usingthe more balanced and sampled average energy of thenetwork and remaining energy of the nodes

(v) In this study the threshold value was better adjustedand CH selection was made faster In this way theenergy of all heterogeneous nodes has been utilizedas much as possible

Wireless Communications and Mobile Computing 3

In this study designing the threshold energy model unlikeheterogeneous algorithms such as other DEEC EDEECEDDEEC HDEEC and TBSE is derived from both a meanand a residual energy of a node in the network In thisstudy the proposed model was compared with DEECEDEEC and EDDEEC as two different simulations usingtheMATLABprogram for network performance throughputof the network energy consumption of the system numberof packets received by BS and average latency Also theproposed model was compared with EH-WSN and ABCOfor alive nodes in the network and energy consumption ofthe system as different simulations All the parameters usedin programming the three methods are identical The resultsshow that the proposed model prolongs the network lifetimeand is better than the other three clustering protocols in termsof energy efficiency

The rest of the paper is listed as follows Section 2 presentsthe basis of the energy efficient model in detail on behalf ofthe three protocols and proposed model Proposed modelis presented in a separate Section 3 Simulation design andcomparison of the results are presented in Section 4 FinallySection 5 concludes the paper in briefly

2 Energy Efficient Modelling

Clustering is importantwhendesigning energy efficientWSNmodels The components of the clustering network structureare explained as follows sensor nodes perform tasks such asdata detection data memory management data routing andprocessing of the data Clusters are the collection units of theWSNs Large sensor networks should be divided into clustersto perform energy efficient WSNs Cluster heads (CHs) arethe leaders of the clusters CHs implement activities can begrouped into data aggregation communication organizationwithin the cluster and communication with the base station(BS) BS sensor node is the point where data obtained fromthe network is collected BS provides communication linkbetween end user and sensor network The end user is aperson accessing the WSN and using the obtained data invarious applications [24]

Figure 1 shows a clustering structure heterogeneousmodel in WSNs

DEEC EDEEC EDDEEC and the proposed model areexplained in detail in this section First we present a two-level heterogeneous model for the DEEC protocol then athree-level heterogeneous network model for both EDEECand EDDEEC protocols and finally the heterogeneous andenergy consumption model of the proposed algorithm

21 DEECModel HeterogeneousWSNs consist of two threeor multiple types sensor nodes in terms of energy levelshardware structure and other special properties [20] TheDEEC protocol is based on a two-level heterogeneous WSNin which the sensor nodes are assumed to have normaland advanced battery levels [11] However multilevel hetero-geneity can be considered for DEEC 1198640 and 1198640119886 representthe initial energy of a normal and advanced sensor noderespectively 119886 indicates how many times energies advancednode has been relative to the normal node The numbers of

Base Station

Normal nodes

Cluster ACluster B

Cluster C Cluster D

Super nodes (CH)

Advanced nodes

Figure 1 A clustering heterogeneousWSNmodel

normal and advanced nodes in the network are 119873119899119898119897 and119873119886119889V119888119889 respectively So total numbers of nodes (119873) in WSNare defined in

119873 = 119873119899119898119897 + 119873119886119889V119888119889 (1)

The total first energy (119864119899119898119897) of the normal nodes in the WSNis given in

119864119899119898119897 = 1198731198991198981198971198640 (2)

The total first energy of the advanced nodes in the WSN(119864119886119889V119888119889) is given in

119864119886119889V119888119889 = 119873119886119889V1198881198891198640119886 (3)

Thus the total first energy of the two-level heterogeneousWSNs is calculated as given in

119864119905119900119905119886119897 = 119864119899119898119897 + 119864119886119889V119888119889 (4)

Heterogeneous WSN becomes homogenous after manyrounds due to the different energy dissipation of the sensornodes CH consumes more energy than sensor nodes andother member nodes After several rounds the energy levelof all sensor nodes changes relative to each other Forthis reason a clustering network protocol that operateswith heterogeneity is more significant than a homogeneousnetwork method [20] Energy consumption of a sensor nodeinvolves models that consume energy so that it can performspecific functions such as sensing processing and wirelesscommunication of collected data [25ndash27] These modelshave become functional by making energy consumption


Table 1 Simulation parameters

Type of parameters Symbol ValueEnergy depletion of the booster to deliver at a shorter distance 119890119891119904 10nJbit1198982Energy depletion of the booster to deliver at a longer distance 119890119886119898119901 00015pJbit1198984Energy depletion of the nodersquos electronics circuit to transmit or receive the signal 119864119890119897119890119888 60nJbitEnergy for data aggregation 119864119863119860 5nJbitsignalThreshold distance 1198890 70mDesired probability of CH 119901119900119901119905 01Total rounds number 119877 5000Data size 119897 5000 bitsNetwork size - 100 x100Sink node position - (5050)Number of sensor nodes 119873 100Normal node numbers 119873119899119898119897 25Advanced node numbers 119873119886119889V119888119889 35Super node numbers 119873119904119906119901119890119903 40Network deployment - Randomly

calculations For the DEEC model it includes the idea ofthe probabilities of the nodes dependent on the start and theremaining energy in addition to the average energy of thenetwork when the CH selection is performed The averageenergy of the network is given as in (5) for 119903 round

119864119886V119892 = 1119873119864119905119900119905119886119897 (1 minus 119903

119877) (5)

As seen in (5) 119864119886V119892 is found as 119864119905119900119905119886119897 is the total energy ofthe 119873 nodes and r round in all rounds 119877 is defined as thenumber of rounds predicted according to the available energyand energy consumed at the current round is given by (6)119864119903119900119906119899119889 refers to the energy consumed for each round

119877 = 119864119905119900119905119886119897119864119903119900119906119899119889 (6)

At the beginning of each round the decision as to whether ornot the nodes are CH is decided by the threshold value Thethreshold value is recommended as in (7) It is important tonote that desired probability (119901119894) is between 0 and 1 which isthe fraction remaining in the inverse of the 119901119894with r This iswhy mod is usedThis residual is subtracted by 1 and 119879(119870119894) iscalculated

119879 (119870119894) =

1199011198941 minus 119901119894 (mod (119903 1119901119894)) 119894119891 119878119894 isin 1198660 119900119905ℎ119890119903119908119894119904119890

(7)

The selection for CH G includes the appropriate set of nodesand 119901119894 is the desired possibility for CH 119878119894 is 119894 node withinthe cluster The possibilities for CH selection in the DEECmodel are given in (8) 119864119894(119903) is the energy of the node 119901119900119901119905is used constant probability for CH In (8) because 119864119886V119892 isrecalculated for each round 119864119886V119892 is important to be here If

it is also assumed to be 119864119894(119903)119901119900119901119905 = 119864119886V119892 then the sum of allpossible states of 119901119894 is 1 This case is also true for (11)

119901119894 =

119864119894 (119903) 119901119900119901119905(1 + 119886) 119864119886V119892 119894119891 119899119900119903119898119886119897 119899119900119889119890119864119894 (119903) 119901119900119901119905119886(1 + 119886) 119864119886V119892 119894119891 119886119889V119886119899119888119890119889 119899119900119889119890

(8)

In this study the DEEC model was designed as threelevels because the equilibrium heterogeneous structure wasconsidered in the simulation comparisons Moreover theprobability of CH selection in a multilevel heterogeneousnetwork model is as in

119901119898119906119897119905119894minus119897119890V119890119897 =119901119900119901119905119873(1 + 119886)(119873 + sum119873119894=1 119886119894)

(9)

119901119900119901119905 is constant and given value of this in Table 1 It only used 119886coefficient as we show the multilevel heterogeneous networkWhen the 119873 119886 is in the case of a denominator 119901119898119906119897119905119894minus119897119890V119890119897 isfound In this case the probability of 119901119898119906119897119905119894minus119897119890V119890119897 is in only onemultiplication factor in119873 nodes

22 EDEEC Model EDEEC uses the idea of three levels ofheterogeneous sensor networks This is different from theDEEC model in this sense The EDEEC model is based ona three-level heterogeneous WSN where sensor nodes arethought to have normal advanced and super-battery levels1198640 1198640119886 and 1198640119887 represent the starting energy of a normaladvanced and super sensor node respectively The numbers119886 and 119887 determine how many times more energy is advancedand super nodes are more than normal nodes [19] Since119873 isthe number of nodes in the network the numbers of normaladvanced and super nodes in the network are 119873119899119898119897 119873119886119889V119888119889and 119873119904119906119901119890119903 respectively Thus the total first energy of thethree-level heterogeneous WSNs is calculated as given in

119864119905119900119905119886119897 = 119864119899119898119897 + 119864119886119889V119888119889 + 119864119904119906119901119890119903 (10)


Transmit electronics TX amplifier Receive

electronicsEtx(ld)

Eeleclowastl eamplowastllowastd

Erx(ld)

Electlowastl

l bitpackets

l bitpackets

d

Figure 2 Energy dissipation model

Also the proposed probabilities for the CH selection for theEDEEC model are given in (11)

The threshold used to select CH in the EDEEC model isproposed as in (7) for any types of nodes

119901119894 =

119864119894 (119903) 119901119900119901119905(1 + 119886 + 119887) 119864119886V119892 119894119891 119899119900119903119898119886119897 119899119900119889119890119864119894 (119903) 119901119900119901119905119886

(1 + 119886 + 119887) 119864119886V119892 119894119891 119886119889V119886119899119888119890119889 119899119900119889119890119864119894 (119903) 119901119900119901119905119887

(1 + 119886 + 119887) 119864119886V119892 119894119891 119904119906119901119890119903 119899119900119889119890

(11)

23 EDDEEC Model DDEEC model utilizes the same net-work structure with other energy models like EDEEC Whena bit of data is sent or received for energy consumption anddistance for the proposed model the energy dissipation ofthe sensor node 119864119879 is calculated as (12) Figure 2 showsthe network energy consumption model 119897 is data size 119864119890119897119890119888refers to the energy consumption per bit of the sensor toelectronically operate the transmitter or receiver 119890119891119904 and119890119886119898119901 denote types of radio amplifiers for free space andmultiple paths respectively 119864119879119883119877119883(119897 119889) refers to the energyconsumption in sending and receiving data for 119897 bits

119864119879 = 119864119879119883119877119883 (119897 119889) =119897119864119890119897119890119888 + 1198971198901198911199041198892 119889 lt 1198890119897119864119890119897119890119888 + 1198971198901198861198981199011198894 119889 ge 1198890

(12)

This method is based on the EDEEC model The modelincludes the idea of the probabilities dependent on the startand the remaining energy of the nodes in addition to theaverage energy of the network when CH is selected Theaverage energy of the network is given as (4) for rround Ris given by (6)

3 Proposed TBSDEEC (Threshold BalancedSampled DEEC) Model

In the proposed model a three-level heterogeneous networkstructure is considered in the same way as the EDDEECmodel The total energy is calculated in the same way asin (10) When a bit of data is sent or received for energyconsumption and distance for the proposed model theenergy dissipation of the sensor node 119864119879 is calculated inthe same way as in (12) The proposed method is basedon the EDDEEC model The model includes the idea ofthe probabilities dependent on the start and the remainingenergy of the nodes in addition to the average energy of

the network when CH is selected The average energy of thenetwork is given as in (5) for 119903 round119877 is given by (6)119864119903119900119906119899119889is the energy consumed in a sensor network during a singleround At the beginning of each round the decision as towhether or not the nodes are CH is decided by the thresholdvalue The threshold value is recommended as in (13) forany type of nodes In this study it is proposed to use theenergy-balanced sampled value(120591119864119904119886119898119901119897119890) for the thresholdvalue (119879(119870119894)) calculation which is the main contribution ofour study Also 119864119904119886119898119901119897119890 depends on i

119879 (119870119894) =

1199011198941 minus 119901119894 (mod (119903 1119901119894))

(120591119864119904119886119898119901119897119890)119891119900119903 119886119899119910 119905119910119901119890119904 119900119891 119899119900119889119890

119894119891 119878119894 isin 119866(13)

120591 is value between 0 and 1 used as a weighted ratio andis used together with the energy-balanced value (119864119904119886119898119901119897119890)In fact it is not possible to know the value of 120591 because ofthe fact that network formation for the next network tour isnot known The appropriate value of 120591 can be determinedby many simulations using random networks In the firstscenario of this study the 120591 value was set at 065 for asimulation In the second scenario 1000 simulations were runfor random networks and average values of all simulationswere obtainedThe energy balance value (119864119904119886119898119901119897119890) for the nextround is given as in

119864119904119886119898119901119897119890 (119894) =1003816100381610038161003816100381610038161003816100381610038161 minus 119864119886V119892

119864119894 (119903)100381610038161003816100381610038161003816100381610038161003816

(14)

The aim of this equation in our study as different from othersis to minimize and balance the energy depletion betweenthe nodes which will increase the stability period andthe sensor network life 119864119904119886119898119901119897119890(119894) is calculated in equation(14) and is determined balanced of the nodes energy loadusing sampled with 1 In this sense it has been sampled119864119904119886119898119901119897119890(119894) by dividing the average energy by the remainingenergy Also if 119864119894(119903) is less than 119864119886V119892 then 119864119904119886119898119901119897119890(119894) willbe negative so we use the absolute form of the expressionin the 119864119904119886119898119901119897119890(119894) If 119864119894(119903) equals 119864119886V119892 then 119864119904119886119898119901119897119890(119894) willbe zero and we consider that 119879(119870119894) has already becomebalanced There is no need to use the (120591119864119904119886119898119901119897119890) expressionin the threshold equation So more precise threshold energy119879(119870119894) was obtained in this study In addition after sometours the super and advanced nodes remain at the sameenergy level as normal nodes In this sense DEEC cannotuse advanced nodes well and cannot manage advanced nodes


Start

Ei(r)gt Tlimit

Compute pi for all node types

Update the pi based on Tlimit Cluster member=inode

CH=inode

End

TlimitltT(Ki)

Yes

No

No

Yes

Figure 3 The flowchart of the working mechanism of proposedmodel

such as EDEEC super nodes well Also EDDEEC may nothave guaranteed a balanced and adapted threshold for CHselection To solve this disadvantage problem in a three-level heterogeneous network and to prevent the super andadvanced nodes from being wasted the changes defined bythe proposed method have been presented in the thresholdfunction This change depends on the constant limit level(119879119897119894119898119894119905) This value indicates that the advanced and supernodes have the equal energy as their normal nodes Fromthis idea it can be understood that under 119879119897119894119898119894119905 all normaladvanced and super nodes are equally likely to be chosen asCH In the proposedmodel the probabilities proposed in CHselection are given in (15) The pseudocode of the proposedmethod and flowchart are given in Algorithm 2 and Figure 3respectively As shown in the line (12) of the algorithm when119879119897119894119898119894119905 is smaller than the 119879(119870119894) i node is chosen as CH Inanother case i node is chosen as cluster member

The value of the energy level 119879119897119894119898119894119905 is calculated as in (16)

119901119894

=

119864119894 (119903) 1198640119901119900119901119905119864119886V119892119864119905119900119905119886119897 for normal nodes (if (119864119894 (119903) gt 119879119897119894119898119894119905)

119864119894 (119903) 1198640119901119900119901119905a119864119886V119892119864119905119900119905119886119897 for advanced nodes (if (119864119894 (119903) gt 119879119897119894119898119894119905)

119864119894 (119903) 1198640119901119900119901119905b119864119886V119892119864119905119900119905119886119897 for super nodes (if (119864119894 (119903) gt 119879119897119894119898119894119905)

(15)

119879119897119894119898119894119905 = 1205911198640 (16)

The pseudocode of the 119879(119870119894) for each round is presented inAlgorithm 1

(1) For r=11MaxRound(2) Compute 119864119894(119903) and 119864119886V119892(3) Compute 119864119904119886119898119901119897119890 according to equation (14)(4) Compute 119901119894 according to equation (15)(5) if (node isin G) then(6) 119879(119870119894) larr997888 119901119894

1 minus 119901119894(mod(119903 1119901119894)) (120591119864119904119886119898119901119897119890)(7) Else then(8) Goto (4)(9) End if(10) End for

Algorithm 1 The pseudocode of the 119879(119870119894)

(1) Compute all alive nodes(2) Compute the CH percentage(3) For r=11MaxRound(4) Compute 119864119894(119903) and 119864119903119900119906119899119889(5) Compute 119864119886V119892 at current round(6) If (119864119894(119903) gt 119879119897119894119898119894119905) then(7) Compute 119901119894 for all node types(8) Else then(9) Update the 119901119894 based on 119879119897119894119898119894119905(10) End if(11) if (node isin G) then(12) if ( 119879119897119894119898119894119905 lt 119879(119870119894) then(13) CHlarr997888 i node(14) Else then(15) Cluster memberlarr997888 i node(16) End if(17) Else then(18) Goto (14)(19) End if(20) End For

Algorithm 2 The proposed algorithm

4 Simulation Results

In this study the simulation results of DEEC EDEECEDDEEC and the proposed protocol for three levels ofheterogeneous WSNs were analysed using MATLAB pro-gramming Two different scenarios are considered in thisstudy While WSN was being constructed 100 sensor nodeswere randomly distributed in a 100m by 100m area with acentrally positioned BS It is estimated that all the sensornodes are in a fixed position and that there is no energyloss owing to the deterioration between the signals of allthe nodes The quality performance criteria utilized foranalysis of models are live nodes in the network number ofpackets received byBS energy consumption throughput andaverage latency

(i) Alive nodes on the network the living nodes metrictakes into account the fact that the first node is starting to dieand all the nodes have died


Table 2 Comparison of algorithms (Scenario 1)

Algorithms First dying rounds of the nodes Dead rounds of the all nodesDEEC 1512 2813EDEEC 2179 3654EDDEEC 2557 4258proposed 2761 4536

100

90

80

70

60

50

40

30

20

10

0100908070605040302010

Figure 4 Cluster formation (Scenario 1)

(ii) The number of packets received by BS with thisperformance metric the total number of packets received byBS is considered

(iii) Energy consumption of methods with this perfor-mance metric the total energy depletion over the lifetime ofthe network is considered

(iv) Throughput this performance metric tells us howaccurately the data is transmitted and how accurate the datais delivered from the first bit to the last bit

(v) Average latency average latency is time delay whendelivering data packets to the BS

Table 1 shows the simulation parameters used in thescenarios

41 Scenario 1 In this scenario a network consisting of25 normal nodes with 1198640 initial energy is assumed in thesensor network There are 35 advanced nodes with 2 moreenergies than that of normal nodes (a = 2) and 40 supernodes with 3 more energies (b = 3) In this scenario theweighted ratio (120591) is 065 Figure 4 depicts sensor nodesclusters and CHs that are randomly distributed after thenetworkrsquos setup phase The model is proposed as shownin Figure 5(b) the numbers of packets received by BS areapproximately 60x 105 625x 105 75x 105 and 825x 105for DEEC EDEEC EDDEEC and the proposed methodrespectively As shown in Figure 5(c) it has performed betterthan the other network-based methods taking into accountthe fact that network throughput drops to approximately2530th round for DEEC 2761th round for EDEEC 4375thround for EDDEEC and 4760th round for the proposedmethod When the algorithm is executed DEEC takes into

consideration the energy of the sensor nodes and the averageenergy of the network EDDEEC considers the remainingenergy of the nodes In the proposed model one nodecontributes to a different threshold balanced energy level Allthese ideas have a significant impact on CH selection criteriaAs shown in Figure 5(d) the DEEC EDEEC EDDEECand suggested routing algorithms completely exhausted theremaining energy of the network at 2410 2605 3365 and3674th rounds respectivelyThroughout the proposedmodellengthens the network lifetime better for longer rounds thanthe three clustering models

Figure 5(a) illustrates the number of live nodes affectingthe network life For DEEC EDEEC EDDEEC and theproposed model the first node died in 1512 2179 2557 and2761th rounds respectively and all nodes died in 2813 36544258 and 4536th rounds respectively (see Table 2) Themodel proposed in as shown in Figure 5(b) demonstrates thatthe number of packets received by BS is more than the otherprotocols As can be seen in Figure 5(e) as the number ofrounds increases average latency reduces for all methods

For example in the 2500th round DEEC EDEECEDDEEC and the proposed algorithm are seen as 800 750640 and 550 milliseconds respectively This means that theproposed algorithm delivers data packets with minimumlatency owing to the fact that the proposed protocol deliverdata packets to the BS with minimum relay after calculatingthe optimal possible distance for the next hop moreoverthe CHs are placed at optimal distance to BS thus theresults obtained from the simulations show quality of thisstudy

42 Scenario 2 The main purpose of Scenario 2 is to bring119879119897119894119898119894119905 in (16) to optimistic value 119879119897119894119898119894119905 how sensitive ismeasured i node is more correctly assigned to the CH asaccurately as expressed in lines (12) and (13) of Algorithm 2In this way this method makes it easier to increase the lifeof the network In this sense in Scenario 1 the value of120591 was assumed to be a constant value of 065 In Scenario2 120591 is found 058 as a result of 1000 simulations Thusaccording to Table 2 as a result of Scenario 1 the first nodein the network died in 2761th round though according toTable 3 the results of Scenario 2 the first node in the networkdied in 2628th round That is the life of the network hasincreased by about 6 As a result energy usage and networklife have had positive results in terms of all algorithms Allother parameters considered for the simulation are as inTable 1 Figure 6 depicts sensor nodes clusters and CHsthat are randomly distributed after the networkrsquos setup phase


100

90

80

70

60

50

40

30

20

10

0

Num

ber o

f aliv

e nod

es

0

500

1000

1500

2000

2500

3000

3500

4000

4500

5000

number of rounds

DEECEDEECEDDEECProposed

(a)

90

75

60

45

30

15

Num

ber o

f pac

kets

rece

ived

by

BS

105

0

500

1000

1500

2000

2500

3000

3500

4000

4500

5000

number of rounds

EDEECDEEC

EDDEECProposed

(b)

900

800

700

600

500

400

300

200

100

0

thro

ughp

ut in

A b

ites

1000

0

500

1000

1500

2000

2500

3000

3500

4000

4500

5000

number of rounds

EDEECDEEC

EDDEECProposed

(c)

20

18

16

14

12

10

8

6

4

2

0

resid

ual e

nerg

y of

the n

etw

ork

(J)

0

500

1000

1500

2000

2500

3000

3500

4000

4500

5000

number of rounds

EDEECDEEC

EDDEECProposed

(d)1000

900

800

700

600

500

400

300

200

100

aver

age l

aten

cy (m

s)

500

1000

1500

2000

2500

3000

3500

4000

4500

5000

number of rounds

EDEECDEEC

EDDEECProposed

(e)

Figure 5 Simulation results of the proposed TBSDEEC model and the other methods (Scenario 1) (a) Number of alive nodes throughnetwork lifetime (b)The number of packets received by BS (c) Network throughput (d) Network energy consumption (e) Average latency

for scenario 2 This scenario is the improved version of thescenario 1 This means we adjust the 120591 value in a fix statusfor achieving longer network lifetime and providing moreperformance on rounds

Figure 7(a) shows the number of living nodes affecting thelongevity of the network For the DEEC EDEEC EDDEECand the recommended protocol the first node died in 15562265 2614 and 2928th rounds respectively and all nodes


Table 3 Comparison of the algorithms (Scenario 2)


100

90

80

70

60

50

40

30

20

10

0100908070605040302010


died in 2876 3756 4305 and 4827th rounds respectively(see Table 3) Figure 7(b) illustrates that the numbers ofpackets received by BS are more than 90 x 105packetswhich is more for the proposed algorithm than the otherprotocols It is also clear that for all algorithms BS hasmore packages than the first scenario In Figure 7(c) theproposed method performed better than the other methodsaccording to network throughput and was found to end afterapproximately 4910th round For all algorithms this scenarioreflects the superiority of the first and has best results Theremaining energy of the network is shown in Figure 7(d) Asseen in Figure 7(d) when the proposed protocol is used itconsumes all the energy of the network after 4920th round Inthis sense the proposed model extends the network lifetimebetter than the three clustering models for longer roundsIn addition all simulation results show energy efficiencyfor longer rounds compared to performance criteria whichmeans that the network lifetime is longer

As can be seen in Figure 7(e) as the number of roundsincreases average latency reduces for all methods For exam-ple in the 2000th round DEEC EDEEC EDDEEC andthe proposed algorithm are seen as 700 610 480 and 400milliseconds respectively

As shown in Table 4 we also analysed the influenceof 119864119894(119903) based on 119879(119870119894) We derived energy consumptiondifferences between proposed and the other algorithms aspercentage using Tables 2 and 3 thanks to dead rounds ofthe all nodes When 120591 and 119879(119870119894) are calculated and fitted to(13) the CH selection is made very often and more nodesbecome CH and energy is consumed in a balanced levelSo energy consumption decreases and the balanced state of119879(119870119894) is taken into account As seen in Table 4 the proposed

Table 4 The performance difference as percentage ()

The other algorithms compared The proposed algorithmFor Scenario 1 For Scenario 2

DEEC 365 392EDEEC 1764 2142EDDEEC 556 1042

algorithm with Scenario 1 enhanced the network lifetime365 1764 and 556 for DEEC EDEEC and EDDEECrespectively Also the proposed algorithm with Scenario 2prolonged the network lifetime 392 2142 and 1042 forDEEC EDEEC and EDDEEC respectively

After that we compared other two newly protocols withour proposedmethod according to alive nodes in the networkand residual energy of the network In this scenario weexecute the simulations based on scenario 1 and Table 1parameters

Figure 8(a) shows the number of living nodes affecting thelongevity of the network For the EH-WSN [21] ABCO [22]and the proposed protocol the first node died in 2560 2653and 2761th rounds respectively and all nodes died in 43254410 and 4536th rounds as seen in Table 5 It is clear that ourmethod has the most performance

As seen in Figure 8(b) when the proposed protocolis used it consumes all the energy of the network after3674th round In this mean because 119879119897119894119898119894119905 and 119879(119870119894) arecalculated in the most accurate and balanced way and totalenergy is used in more balanced and distributed way whileCH is chosen the proposed method has shown superiorperformance compared to other existing methods In thissense the proposed model extends the network lifetimebetter than the two clustering models for longer roundsIn addition all simulation results show energy efficiencyfor longer rounds compared to performance criteria whichmeans that the network lifetime is longer

5 Conclusions

This study presents an energy efficient clustering heteroge-neous protocol based DEEC variants in distributed WSNsWe analysed the performances of the proposed protocol withdifferent two scenarios in comparison with DEEC EDEECEDDEEC EH-WSN and ABCO protocols in terms of cri-teria alive nodes during the network life and throughputof the sensor network number of packets received by BSin the network energy depletion and average latency ofthe algorithms in MATLAB simulation environment The


100

90

80

70

60

50

40

30

20

10

0

Num

ber o

f aliv

e nod

es

0

500

1000

1500

2000

2500

3000

3500

4000

4500

5000

number of rounds


(a)

75

60

45

30

15

0

Num

ber o

f pac

kets

rece

ived

by

BS

105

0

500

1000

1500

2000

2500

3000

3500

4000

4500

5000

number of rounds

EDEECDEEC

EDDEECProposed

105

90

(b)

900

800

700

600

500

400

300

200

100

0

thro

ughp

ut in

A b

ites

1000

0

500

1000

1500

2000

2500

3000

3500

4000

4500

5000

number of rounds

EDEECDEEC

EDDEECProposed

(c)

20

18

16

14

12

10

8

6

4

2

0

resid

ual e

nerg

y of

the n

etw

ork

(J)

0

500

1000

1500

2000

2500

3000

3500

4000

4500

5000

number of rounds

EDEECDEEC

EDDEECProposed

(d)

1000

900

800

700

600

500

400

300

200

100

aver

age l

aten

cy (m

s)

500

1000

1500

2000

2500

3000

3500

4000

4500

5000

number of rounds

EDEECDEEC

EDDEECProposed

(e)


proposed method (TBSDEEC) demonstrates its superiorityover the other methods in terms of the parameters concernedand has been found to be advantageous with respect toenergy consumption and threshold balanced sampled value

considered Also the consumption of the first scenario asa result of the best balanced value obtained and prolongedthe lifetime of the network In this sense we contribute adifferent energy consumption CH selection method to the


Table 5 Comparison of the other algorithms

Algorithms First dying rounds of the nodes Dead rounds of the all nodesEH-WSN [21] 2560 4325ABCO [22] 2653 4410proposed 2761 4536

100

90

80

70

60

50

40

30

20

10

0

num

ber o

f aliv

e nod

es

0

500

1000

1500

2000

2500

3000

3500

4000

4500

5000

number of rounds(a)

20

18

16

14

12

10

8

6

4

2

0

resid

ual e

nerg

y of

the n

etw

ork

(J)

0

500

1000

1500

2000

2500

3000

3500

4000

4500

5000

number of rounds(b)

Figure 8 Other simulation results (a) Number of alive nodes through network lifetime (b) Network energy consumption

heterogeneousWSNs and the proposed algorithm can inspireother researchers in the future works

Data Availability

The data used to support the findings of this study areincluded within the article

Conflicts of Interest

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

References

[1] S Vancin and E Erdem ldquoImplementation of the vehiclerecognition systems using wireless magnetic sensorsrdquo Sadhanavol 42 no 6 pp 841ndash854 2017

[2] J Burrell T Brooke and R Beckwith ldquoVineyard ComputingSensor Networks in Agricultural Productionrdquo InternationalJournal of Computer Applications (0975 ndash 8887) vol 97 no 7pp 9ndash16 2014

[3] R Kumar and R Kaur ldquoEvaluating the Performance of DEECVariantsrdquo International Journal of Computer Applications vol97 no 7 pp 9ndash16 2014

[4] W R Heinzelman A Chandrakasan and H Balakrish-nan ldquoEnergy-efficient communication protocol for wirelessmicrosensor networksrdquo in Proceedings of the 33rd AnnualHawaii International Conference on System Siences (HICSS rsquo00)vol 2 IEEE January 2000

[5] T Kang J Yun H Lee et al ldquoA Clustering Method for EnergyEfficient Routing in Wireless Sensor Networksrdquo in Proceedingsof the International Conference on Electronics Hardware Wire-less and Optical Communications pp 133ndash138 Corfu IslandGreece 2007

[6] J Singh B P Singh and S Shaw ldquoA new LEACH-based routingprotocol for energy optimization in Wireless Sensor Networkrdquoin Proceedings of the 5th IEEE International Conference onComputer and Communication Technology ICCCT 2014 pp181ndash186 September 2014

[7] A Kumar and V Kumar Katiyar ldquoIntelligent Cluster RoutingAn Energy Efficient Approach for Routing in Wireless SensorNetworksrdquo International Journal of Computer Applications vol110 no 5 pp 18ndash22 2015

[8] S Lindsey and C S Raghavendra ldquoPEGASIS power-efficientgathering in sensor information systemsrdquo in Proceedings of theIEEE Aerospace Conference vol 3 pp 1ndash6 Big Sky MT USAMarch 2002

[9] N AT and D SM ldquoA New Energy Efficient Clustering-basedProtocol for HeterogeneousWireless Sensor Networksrdquo Journalof Electrical amp Electronic Systems vol 04 no 03 2015

[10] J-Y Lee K-D Jung and D Lee ldquoThe routing technology ofwireless sensor networks using the stochastic cluster head selec-tion methodrdquo International Journal of Control and Automationvol 8 no 7 pp 385ndash394 2015

[11] M Aslam E U Munir M M Rafique and X Hu ldquoAdaptiveenergy-efficient clustering path planning routing protocols forheterogeneous wireless sensor networksrdquo Sustainable Comput-ing vol 12 pp 57ndash71 2016

[12] V M Mohammad and M Noorian ldquoAn Efficient Data Aggre-gation Method in Wireless Sensor Network based on the SVDrdquoin Proceedings of theThe International Conference on Computing


Technology and Information Management (ICCTIM) Society ofDigital Information andWireless Communication 2014

[13] S Dhankhar and E S Singh ldquoPerformance Comparisonof LEACH ampamp HEED Clustering Protocols in WSNusing MATLAB-A Reviewrdquo International Journal of TechnicalResearch vol 5 no 1 pp 167ndash170 2016

[14] K H Krishna Y S Babu and T Kumar ldquoWireless NetworkTopological Routing in Wireless Sensor Networksrdquo ProcediaComputer Science vol 79 pp 817ndash826 2016

[15] H P Gupta S V Rao and T Venkatesh ldquoAnalysis of stochasticcoverage and connectivity in three-dimensional heterogeneousdirectional wireless sensor networksrdquo Pervasive and MobileComputing vol 29 pp 38ndash56 2016

[16] G Smaragdakis I Matta and A Bestavros ldquoSEP A StableElection Protocol for clustered heterogeneous wireless sensornetworkrdquo in Proceedings of the Second International Work-shop on Sensor and Actor Network Protocols and Applications(SANPA) vol 97 pp 1ndash11 2004

[17] L Qing Q Zhu andMWang ldquoDesign of a distributed energy-efficient clustering algorithm for heterogeneous wireless sensornetworksrdquoComputer Communications vol 29 no 12 pp 2230ndash2237 2006

[18] B Elbhiri S Rachid S El Fkihi and D Aboutajdine ldquoDevel-oped Distributed Energy-Efficient Clustering (DDEEC) forheterogeneous wireless sensor networksrdquo in Proceedings of the2010 5th International Symposium on IV Communications andMobile Networks ISIVC pp 1ndash4 Morocco October 2010

[19] P Saini and A K Sharma ldquoE-DEEC - Enhanced distributedenergy efficient clustering scheme for heterogeneous WSNrdquo inProceedings of the 2010 1st International Conference on ParallelDistributed and Grid Computing PDGC - 2010 pp 914ndash919October 2010

[20] N Javaid T N Qureshi A H Khan A Iqbal E Akhtarand M Ishfaq ldquoEDDEEC Enhanced Developed DistributedEnergy-Efficient Clustering for HeterogeneousWireless SensorNetworksrdquoProcedia Computer Science vol 19 pp 914ndash919 2013

[21] S M Bozorgi A Shokouhi Rostami A A R Hosseinabadi andV E Balas ldquoA new clustering protocol for energy harvesting-wireless sensor networksrdquo Computers and Electrical Engineer-ing vol 64 pp 233ndash247 2017

[22] A Gambhir A Payal and R Arya ldquoPerformance analysis ofartificial bee colony optimization based clustering protocol invarious scenarios of WSNrdquo Procedia Computer Science vol 132pp 183ndash188 2018

[23] M K Khan M Shiraz K Z Ghafoor S Khan A S Sadiq andGAhmed ldquoEE-MRP Energy-EfficientMultistage Routing Pro-tocol for Wireless Sensor Networksrdquo Wireless Communicationsand Mobile Computing vol 2018 2018

[24] S Vancin and E Erdem ldquoPerformance analysis of the energyefficient clustering models in wireless sensor networksrdquo inProceedings of the 2017 24th IEEE International Conference onElectronics Circuits and Systems (ICECS) pp 247ndash251 BatumiDecember 2017

[25] F Engmann F A Katsriku J Abdulai K S Adu-Manu andF K Banaseka ldquoProlonging the Lifetime of Wireless SensorNetworks A Review of Current TechniquesrdquoWireless Commu-nications and Mobile Computing vol 2018 pp 1ndash23 2018

[26] M Nicolae D Popescu R Dobrescu and I Costea ldquoSchedul-ing mechanism for energy-efficient communication in hybridwireless sensor networksrdquo Control Engineering and AppliedInformatics vol 18 no 2 pp 95ndash102 2016

[27] A A Babayo M H Anisi and I Ali ldquoA Review on energymanagement schemes in energy harvesting wireless sensornetworksrdquoRenewable amp Sustainable Energy Reviews vol 76 pp1176ndash1184 2017

International Journal of

AerospaceEngineeringHindawiwwwhindawicom Volume 2018

RoboticsJournal of

Hindawiwwwhindawicom Volume 2018


Active and Passive Electronic Components

VLSI Design



Shock and Vibration


Civil EngineeringAdvances in

Acoustics and VibrationAdvances in



Electrical and Computer Engineering

Journal of

Advances inOptoElectronics

Hindawiwwwhindawicom

Volume 2018

Hindawi Publishing Corporation httpwwwhindawicom Volume 2013Hindawiwwwhindawicom

The Scientific World Journal

Volume 2018

Control Scienceand Engineering

Journal of



Journal ofEngineeringVolume 2018

SensorsJournal of



RotatingMachinery


Modelling ampSimulationin EngineeringHindawiwwwhindawicom Volume 2018


Chemical EngineeringInternational Journal of Antennas and

Propagation




Navigation and Observation


Hindawi

wwwhindawicom Volume 2018

Advances in

Multimedia

Submit your manuscripts atwwwhindawicom


routing protocol based on LEACH for energy optimizationis proposed It is understood that this algorithm is moreefficient than the LEACH algorithmby selecting cluster headsequally The paper [6] presents a modified LEACH derivedfrom the LEACH algorithm In [7] a mobile sink improvedenergy efficient algorithm is presented and compared withmod-LEACH and PEGASIS [8] In [9] a new energy efficient(EE) clustering based method is proposed for single passheterogeneousWSNs Simulations inMATLAB show that thementioned method has a 162-189 times better stability thanknown protocols such as LEACHDEEC and SEP In [10] thestability of the cluster is reduced because the LEACHprotocolon an irregular network causes a decrease in aggregate dataefficiency For this reason this article [10] suggests a methodof selecting a cluster head to improve the LEACH protocol inorder to increase cluster head stability For this purpose anLEACH variant combined with HEED and LEACH protocolis proposed and this method is approved by simulation In[11] two energy efficient route planning routing protocols areproposed for three levels of heterogeneous WSNs namelyCentral Energy Efficiency Clustering (CEEC) with Two-Hop Heterogeneity awareness (THCEEC) and AdvancedEqualization (ACEEC) Comprehensive simulation resultshave provided CEEC ACEEC and THCEEC central clusterdeployments with improved reliability and energy efficiencyperformance providing better network lifetime and success-ful data transmission than LEEC SEP ESEP and DEECrsquostraditional distributed routing protocols In addition ACEECperforms CEEC and provides more network stability timeAnalytical evaluation shows that THCEEC performs CEECACEEC and other existing road planning routing protocolsThe study [12] suggests an efficient method of collecting datawith a support vector in the WSN In [13] performanceevaluation of clustering protocols is presented in WSNsClustering of sensor nodes is an effective technique inreaching these targets With this technique other clusteringmodels (LEACH LEACH-C and HEED) were evaluated andcompared At the end of these clustering methods are com-pared with depending on several criteria such as convergencespeed cluster stability cluster overlap location awarenessand node mobility support In another study [14] the studyof various routing models for sensor networks offers a surveywith classification on behalf of kinds of models The threemain categories examined are data-centric hierarchical andlocation-based Routing methods and algorithms each havea common purpose to better output and extend the usefullife of the sensor network A comparison was made betweenflood and direct diffusion two routing protocols based onnetwork throughput and lifetime Simulation of AODV (AdHoc onDemand Distance Vector) was also performed on twotopologies with the same source and target nodes The study[15] presents random coverage and connectivity analysis inheterogeneous WSNs with three-dimensions The study [16]suggests that the SEP algorithm in which each sensor nodein a two-level heterogeneous sensor network independentlyidentifies itself as a CH on the basis of the first energyrelative to the other sensor nodes of the sensor networkThe study [17] presents a method named DEEC by whichthe CH selection is considered to depend on the ratio of

the remaining energy of the node and the average energyof the sensor network In a study [18] the DDEEC protocolis presented based on the recalibration of the energy forCH This protocol has been optimized by the public wirelessnetwork In this sense it is more likely that advanced nodeswill be chosen as CH in the first broadcast rounds Also whenenergy is reduced these sensor nodes will have the sameprobability of CH selection as normal sensor nodesThe study[19] demonstrates EDEEC a clustering method with a three-level heterogeneous structure that yields a high amount ofenergy level called super sensor nodes In one study [20] aclustering protocol known as EDDEEC was introduced CHselection probability is dependent on the remaining energyquality of the sensor nodes with the average energy of theWSN In one study the opportunity of each node to be chosenas a CH is determined towards to its energy level and to theamount of depleted energy Nodes with higher opportunityhave less delay times The node with the smallest time delaycomparing to its neighbours is chosen as CH After choosinga CH and forming a cluster all nodes of each cluster beginto send packets to the CH depend on energy-aware multihoprouting Then CH sends these packets to the BS by means ofmultihop routing [21] The aim of paper [22] is to analyse theperformance of artificial bee colony optimization algorithm(ABCO) depending on clusteringmethod utilized to improvethe network lifetime The node with highest energy in acluster is chosen as CH in a determined period and the allfield is reclustered depending on the selected CHs In onepaper [23] an energy efficient routing protocol for wirelesssensor networks is recommended This method consists of arouting algorithm for the transmission of data CH choosingalgorithm and a scheme for the formation of clusters

13 Contributions and Motivations Our main contributionscan be listed as follows

(i) The proposed method (TBSDEEC) is a DEEC-basedenergy efficient three-level heterogeneous clusteringmodel named TBSDEEC for distributed sensor net-works

(ii) The proposed method uses an EDDEEC-like net-workmodel However while the energy consumptionmodel is designing the threshold energy level differsfrom other protocols

(iii) In this study we describe a balanced value calledthreshold balanced sample energy (TBSE) which wasproposed and contributed to the calculation of thethreshold value

(iv) Contrary to other studies the model we proposedprovides more accurate and precise solutions for theselection of the CH and the threshold value usingthe more balanced and sampled average energy of thenetwork and remaining energy of the nodes

(v) In this study the threshold value was better adjustedand CH selection was made faster In this way theenergy of all heterogeneous nodes has been utilizedas much as possible









Base Station

Normal nodes

Cluster ACluster B

Cluster C Cluster D

Super nodes (CH)

Advanced nodes



119873 = 119873119899119898119897 + 119873119886119889V119888119889 (1)


119864119899119898119897 = 1198731198991198981198971198640 (2)


119864119886119889V119888119889 = 119873119886119889V1198881198891198640119886 (3)


119864119905119900119905119886119897 = 119864119899119898119897 + 119864119886119889V119888119889 (4)






119864119886V119892 = 1119873119864119905119900119905119886119897 (1 minus 119903

119877) (5)


119877 = 119864119905119900119905119886119897119864119903119900119906119899119889 (6)


119879 (119870119894) =

1199011198941 minus 119901119894 (mod (119903 1119901119894)) 119894119891 119878119894 isin 1198660 119900119905ℎ119890119903119908119894119904119890

(7)



119901119894 =

119864119894 (119903) 119901119900119901119905(1 + 119886) 119864119886V119892 119894119891 119899119900119903119898119886119897 119899119900119889119890119864119894 (119903) 119901119900119901119905119886(1 + 119886) 119864119886V119892 119894119891 119886119889V119886119899119888119890119889 119899119900119889119890

(8)


119901119898119906119897119905119894minus119897119890V119890119897 =119901119900119901119905119873(1 + 119886)(119873 + sum119873119894=1 119886119894)

(9)



119864119905119900119905119886119897 = 119864119899119898119897 + 119864119886119889V119888119889 + 119864119904119906119901119890119903 (10)



electronicsEtx(ld)


Erx(ld)

Electlowastl

l bitpackets

l bitpackets

d




119901119894 =

119864119894 (119903) 119901119900119901119905(1 + 119886 + 119887) 119864119886V119892 119894119891 119899119900119903119898119886119897 119899119900119889119890119864119894 (119903) 119901119900119901119905119886

(1 + 119886 + 119887) 119864119886V119892 119894119891 119886119889V119886119899119888119890119889 119899119900119889119890119864119894 (119903) 119901119900119901119905119887

(1 + 119886 + 119887) 119864119886V119892 119894119891 119904119906119901119890119903 119899119900119889119890

(11)


119864119879 = 119864119879119883119877119883 (119897 119889) =119897119864119890119897119890119888 + 1198971198901198911199041198892 119889 lt 1198890119897119864119890119897119890119888 + 1198971198901198861198981199011198894 119889 ge 1198890

(12)





119879 (119870119894) =

1199011198941 minus 119901119894 (mod (119903 1119901119894))

(120591119864119904119886119898119901119897119890)119891119900119903 119886119899119910 119905119910119901119890119904 119900119891 119899119900119889119890

119894119891 119878119894 isin 119866(13)


119864119904119886119898119901119897119890 (119894) =1003816100381610038161003816100381610038161003816100381610038161 minus 119864119886V119892

119864119894 (119903)100381610038161003816100381610038161003816100381610038161003816

(14)



Start

Ei(r)gt Tlimit



CH=inode

End

TlimitltT(Ki)

Yes

No

No

Yes




119901119894

=




(15)

119879119897119894119898119894119905 = 1205911198640 (16)













100

90

80

70

60

50

40

30

20

10

0100908070605040302010













100

90

80

70

60

50

40

30

20

10

0

Num

ber o

f aliv

e nod

es

0

500

1000

1500

2000

2500

3000

3500

4000

4500

5000

number of rounds


(a)

90

75

60

45

30

15

Num

ber o

f pac

kets

rece

ived

by

BS

105

0

500

1000

1500

2000

2500

3000

3500

4000

4500

5000

number of rounds

EDEECDEEC

EDDEECProposed

(b)

900

800

700

600

500

400

300

200

100

0

thro

ughp

ut in

A b

ites

1000

0

500

1000

1500

2000

2500

3000

3500

4000

4500

5000

number of rounds

EDEECDEEC

EDDEECProposed

(c)

20

18

16

14

12

10

8

6

4

2

0

resid

ual e

nerg

y of

the n

etw

ork

(J)

0

500

1000

1500

2000

2500

3000

3500

4000

4500

5000

number of rounds

EDEECDEEC

EDDEECProposed

(d)1000

900

800

700

600

500

400

300

200

100

aver

age l

aten

cy (m

s)

500

1000

1500

2000

2500

3000

3500

4000

4500

5000

number of rounds

EDEECDEEC

EDDEECProposed

(e)







100

90

80

70

60

50

40

30

20

10

0100908070605040302010












5 Conclusions



100

90

80

70

60

50

40

30

20

10

0

Num

ber o

f aliv

e nod

es

0

500

1000

1500

2000

2500

3000

3500

4000

4500

5000

number of rounds


(a)

75

60

45

30

15

0

Num

ber o

f pac

kets

rece

ived

by

BS

105

0

500

1000

1500

2000

2500

3000

3500

4000

4500

5000

number of rounds

EDEECDEEC

EDDEECProposed

105

90

(b)

900

800

700

600

500

400

300

200

100

0

thro

ughp

ut in

A b

ites

1000

0

500

1000

1500

2000

2500

3000

3500

4000

4500

5000

number of rounds

EDEECDEEC

EDDEECProposed

(c)

20

18

16

14

12

10

8

6

4

2

0

resid

ual e

nerg

y of

the n

etw

ork

(J)

0

500

1000

1500

2000

2500

3000

3500

4000

4500

5000

number of rounds

EDEECDEEC

EDDEECProposed

(d)

1000

900

800

700

600

500

400

300

200

100

aver

age l

aten

cy (m

s)

500

1000

1500

2000

2500

3000

3500

4000

4500

5000

number of rounds

EDEECDEEC

EDDEECProposed

(e)







100

90

80

70

60

50

40

30

20

10

0

num

ber o

f aliv

e nod

es

0

500

1000

1500

2000

2500

3000

3500

4000

4500

5000

number of rounds(a)

20

18

16

14

12

10

8

6

4

2

0

resid

ual e

nerg

y of

the n

etw

ork

(J)

0

500

1000

1500

2000

2500

3000

3500

4000

4500

5000

number of rounds(b)



Data Availability




References
































RoboticsJournal of




VLSI Design



Shock and Vibration







Journal of



Volume 2018



Volume 2018


Journal of




SensorsJournal of



RotatingMachinery





Propagation






Hindawi


Advances in

Multimedia










Base Station

Normal nodes

Cluster ACluster B

Cluster C Cluster D

Super nodes (CH)

Advanced nodes



119873 = 119873119899119898119897 + 119873119886119889V119888119889 (1)


119864119899119898119897 = 1198731198991198981198971198640 (2)


119864119886119889V119888119889 = 119873119886119889V1198881198891198640119886 (3)


119864119905119900119905119886119897 = 119864119899119898119897 + 119864119886119889V119888119889 (4)






119864119886V119892 = 1119873119864119905119900119905119886119897 (1 minus 119903

119877) (5)


119877 = 119864119905119900119905119886119897119864119903119900119906119899119889 (6)


119879 (119870119894) =

1199011198941 minus 119901119894 (mod (119903 1119901119894)) 119894119891 119878119894 isin 1198660 119900119905ℎ119890119903119908119894119904119890

(7)



119901119894 =

119864119894 (119903) 119901119900119901119905(1 + 119886) 119864119886V119892 119894119891 119899119900119903119898119886119897 119899119900119889119890119864119894 (119903) 119901119900119901119905119886(1 + 119886) 119864119886V119892 119894119891 119886119889V119886119899119888119890119889 119899119900119889119890

(8)


119901119898119906119897119905119894minus119897119890V119890119897 =119901119900119901119905119873(1 + 119886)(119873 + sum119873119894=1 119886119894)

(9)



119864119905119900119905119886119897 = 119864119899119898119897 + 119864119886119889V119888119889 + 119864119904119906119901119890119903 (10)



electronicsEtx(ld)


Erx(ld)

Electlowastl

l bitpackets

l bitpackets

d




119901119894 =

119864119894 (119903) 119901119900119901119905(1 + 119886 + 119887) 119864119886V119892 119894119891 119899119900119903119898119886119897 119899119900119889119890119864119894 (119903) 119901119900119901119905119886

(1 + 119886 + 119887) 119864119886V119892 119894119891 119886119889V119886119899119888119890119889 119899119900119889119890119864119894 (119903) 119901119900119901119905119887

(1 + 119886 + 119887) 119864119886V119892 119894119891 119904119906119901119890119903 119899119900119889119890

(11)


119864119879 = 119864119879119883119877119883 (119897 119889) =119897119864119890119897119890119888 + 1198971198901198911199041198892 119889 lt 1198890119897119864119890119897119890119888 + 1198971198901198861198981199011198894 119889 ge 1198890

(12)





119879 (119870119894) =

1199011198941 minus 119901119894 (mod (119903 1119901119894))

(120591119864119904119886119898119901119897119890)119891119900119903 119886119899119910 119905119910119901119890119904 119900119891 119899119900119889119890

119894119891 119878119894 isin 119866(13)


119864119904119886119898119901119897119890 (119894) =1003816100381610038161003816100381610038161003816100381610038161 minus 119864119886V119892

119864119894 (119903)100381610038161003816100381610038161003816100381610038161003816

(14)



Start

Ei(r)gt Tlimit



CH=inode

End

TlimitltT(Ki)

Yes

No

No

Yes




119901119894

=




(15)

119879119897119894119898119894119905 = 1205911198640 (16)













100

90

80

70

60

50

40

30

20

10

0100908070605040302010













100

90

80

70

60

50

40

30

20

10

0

Num

ber o

f aliv

e nod

es

0

500

1000

1500

2000

2500

3000

3500

4000

4500

5000

number of rounds


(a)

90

75

60

45

30

15

Num

ber o

f pac

kets

rece

ived

by

BS

105

0

500

1000

1500

2000

2500

3000

3500

4000

4500

5000

number of rounds

EDEECDEEC

EDDEECProposed

(b)

900

800

700

600

500

400

300

200

100

0

thro

ughp

ut in

A b

ites

1000

0

500

1000

1500

2000

2500

3000

3500

4000

4500

5000

number of rounds

EDEECDEEC

EDDEECProposed

(c)

20

18

16

14

12

10

8

6

4

2

0

resid

ual e

nerg

y of

the n

etw

ork

(J)

0

500

1000

1500

2000

2500

3000

3500

4000

4500

5000

number of rounds

EDEECDEEC

EDDEECProposed

(d)1000

900

800

700

600

500

400

300

200

100

aver

age l

aten

cy (m

s)

500

1000

1500

2000

2500

3000

3500

4000

4500

5000

number of rounds

EDEECDEEC

EDDEECProposed

(e)







100

90

80

70

60

50

40

30

20

10

0100908070605040302010












5 Conclusions



100

90

80

70

60

50

40

30

20

10

0

Num

ber o

f aliv

e nod

es

0

500

1000

1500

2000

2500

3000

3500

4000

4500

5000

number of rounds


(a)

75

60

45

30

15

0

Num

ber o

f pac

kets

rece

ived

by

BS

105

0

500

1000

1500

2000

2500

3000

3500

4000

4500

5000

number of rounds

EDEECDEEC

EDDEECProposed

105

90

(b)

900

800

700

600

500

400

300

200

100

0

thro

ughp

ut in

A b

ites

1000

0

500

1000

1500

2000

2500

3000

3500

4000

4500

5000

number of rounds

EDEECDEEC

EDDEECProposed

(c)

20

18

16

14

12

10

8

6

4

2

0

resid

ual e

nerg

y of

the n

etw

ork

(J)

0

500

1000

1500

2000

2500

3000

3500

4000

4500

5000

number of rounds

EDEECDEEC

EDDEECProposed

(d)

1000

900

800

700

600

500

400

300

200

100

aver

age l

aten

cy (m

s)

500

1000

1500

2000

2500

3000

3500

4000

4500

5000

number of rounds

EDEECDEEC

EDDEECProposed

(e)







100

90

80

70

60

50

40

30

20

10

0

num

ber o

f aliv

e nod

es

0

500

1000

1500

2000

2500

3000

3500

4000

4500

5000

number of rounds(a)

20

18

16

14

12

10

8

6

4

2

0

resid

ual e

nerg

y of

the n

etw

ork

(J)

0

500

1000

1500

2000

2500

3000

3500

4000

4500

5000

number of rounds(b)



Data Availability




References
































RoboticsJournal of




VLSI Design



Shock and Vibration







Journal of



Volume 2018



Volume 2018


Journal of




SensorsJournal of



RotatingMachinery





Propagation






Hindawi


Advances in

Multimedia






119864119886V119892 = 1119873119864119905119900119905119886119897 (1 minus 119903

119877) (5)


119877 = 119864119905119900119905119886119897119864119903119900119906119899119889 (6)


119879 (119870119894) =

1199011198941 minus 119901119894 (mod (119903 1119901119894)) 119894119891 119878119894 isin 1198660 119900119905ℎ119890119903119908119894119904119890

(7)



119901119894 =

119864119894 (119903) 119901119900119901119905(1 + 119886) 119864119886V119892 119894119891 119899119900119903119898119886119897 119899119900119889119890119864119894 (119903) 119901119900119901119905119886(1 + 119886) 119864119886V119892 119894119891 119886119889V119886119899119888119890119889 119899119900119889119890

(8)


119901119898119906119897119905119894minus119897119890V119890119897 =119901119900119901119905119873(1 + 119886)(119873 + sum119873119894=1 119886119894)

(9)



119864119905119900119905119886119897 = 119864119899119898119897 + 119864119886119889V119888119889 + 119864119904119906119901119890119903 (10)



electronicsEtx(ld)


Erx(ld)

Electlowastl

l bitpackets

l bitpackets

d




119901119894 =

119864119894 (119903) 119901119900119901119905(1 + 119886 + 119887) 119864119886V119892 119894119891 119899119900119903119898119886119897 119899119900119889119890119864119894 (119903) 119901119900119901119905119886

(1 + 119886 + 119887) 119864119886V119892 119894119891 119886119889V119886119899119888119890119889 119899119900119889119890119864119894 (119903) 119901119900119901119905119887

(1 + 119886 + 119887) 119864119886V119892 119894119891 119904119906119901119890119903 119899119900119889119890

(11)


119864119879 = 119864119879119883119877119883 (119897 119889) =119897119864119890119897119890119888 + 1198971198901198911199041198892 119889 lt 1198890119897119864119890119897119890119888 + 1198971198901198861198981199011198894 119889 ge 1198890

(12)





119879 (119870119894) =

1199011198941 minus 119901119894 (mod (119903 1119901119894))

(120591119864119904119886119898119901119897119890)119891119900119903 119886119899119910 119905119910119901119890119904 119900119891 119899119900119889119890

119894119891 119878119894 isin 119866(13)


119864119904119886119898119901119897119890 (119894) =1003816100381610038161003816100381610038161003816100381610038161 minus 119864119886V119892

119864119894 (119903)100381610038161003816100381610038161003816100381610038161003816

(14)



Start

Ei(r)gt Tlimit



CH=inode

End

TlimitltT(Ki)

Yes

No

No

Yes




119901119894

=




(15)

119879119897119894119898119894119905 = 1205911198640 (16)













100

90

80

70

60

50

40

30

20

10

0100908070605040302010













100

90

80

70

60

50

40

30

20

10

0

Num

ber o

f aliv

e nod

es

0

500

1000

1500

2000

2500

3000

3500

4000

4500

5000

number of rounds


(a)

90

75

60

45

30

15

Num

ber o

f pac

kets

rece

ived

by

BS

105

0

500

1000

1500

2000

2500

3000

3500

4000

4500

5000

number of rounds

EDEECDEEC

EDDEECProposed

(b)

900

800

700

600

500

400

300

200

100

0

thro

ughp

ut in

A b

ites

1000

0

500

1000

1500

2000

2500

3000

3500

4000

4500

5000

number of rounds

EDEECDEEC

EDDEECProposed

(c)

20

18

16

14

12

10

8

6

4

2

0

resid

ual e

nerg

y of

the n

etw

ork

(J)

0

500

1000

1500

2000

2500

3000

3500

4000

4500

5000

number of rounds

EDEECDEEC

EDDEECProposed

(d)1000

900

800

700

600

500

400

300

200

100

aver

age l

aten

cy (m

s)

500

1000

1500

2000

2500

3000

3500

4000

4500

5000

number of rounds

EDEECDEEC

EDDEECProposed

(e)







100

90

80

70

60

50

40

30

20

10

0100908070605040302010












5 Conclusions



100

90

80

70

60

50

40

30

20

10

0

Num

ber o

f aliv

e nod

es

0

500

1000

1500

2000

2500

3000

3500

4000

4500

5000

number of rounds


(a)

75

60

45

30

15

0

Num

ber o

f pac

kets

rece

ived

by

BS

105

0

500

1000

1500

2000

2500

3000

3500

4000

4500

5000

number of rounds

EDEECDEEC

EDDEECProposed

105

90

(b)

900

800

700

600

500

400

300

200

100

0

thro

ughp

ut in

A b

ites

1000

0

500

1000

1500

2000

2500

3000

3500

4000

4500

5000

number of rounds

EDEECDEEC

EDDEECProposed

(c)

20

18

16

14

12

10

8

6

4

2

0

resid

ual e

nerg

y of

the n

etw

ork

(J)

0

500

1000

1500

2000

2500

3000

3500

4000

4500

5000

number of rounds

EDEECDEEC

EDDEECProposed

(d)

1000

900

800

700

600

500

400

300

200

100

aver

age l

aten

cy (m

s)

500

1000

1500

2000

2500

3000

3500

4000

4500

5000

number of rounds

EDEECDEEC

EDDEECProposed

(e)







100

90

80

70

60

50

40

30

20

10

0

num

ber o

f aliv

e nod

es

0

500

1000

1500

2000

2500

3000

3500

4000

4500

5000

number of rounds(a)

20

18

16

14

12

10

8

6

4

2

0

resid

ual e

nerg

y of

the n

etw

ork

(J)

0

500

1000

1500

2000

2500

3000

3500

4000

4500

5000

number of rounds(b)



Data Availability




References
































RoboticsJournal of




VLSI Design



Shock and Vibration







Journal of



Volume 2018



Volume 2018


Journal of




SensorsJournal of



RotatingMachinery





Propagation






Hindawi


Advances in

Multimedia




electronicsEtx(ld)


Erx(ld)

Electlowastl

l bitpackets

l bitpackets

d




119901119894 =

119864119894 (119903) 119901119900119901119905(1 + 119886 + 119887) 119864119886V119892 119894119891 119899119900119903119898119886119897 119899119900119889119890119864119894 (119903) 119901119900119901119905119886

(1 + 119886 + 119887) 119864119886V119892 119894119891 119886119889V119886119899119888119890119889 119899119900119889119890119864119894 (119903) 119901119900119901119905119887

(1 + 119886 + 119887) 119864119886V119892 119894119891 119904119906119901119890119903 119899119900119889119890

(11)


119864119879 = 119864119879119883119877119883 (119897 119889) =119897119864119890119897119890119888 + 1198971198901198911199041198892 119889 lt 1198890119897119864119890119897119890119888 + 1198971198901198861198981199011198894 119889 ge 1198890

(12)





119879 (119870119894) =

1199011198941 minus 119901119894 (mod (119903 1119901119894))

(120591119864119904119886119898119901119897119890)119891119900119903 119886119899119910 119905119910119901119890119904 119900119891 119899119900119889119890

119894119891 119878119894 isin 119866(13)


119864119904119886119898119901119897119890 (119894) =1003816100381610038161003816100381610038161003816100381610038161 minus 119864119886V119892

119864119894 (119903)100381610038161003816100381610038161003816100381610038161003816

(14)



Start

Ei(r)gt Tlimit



CH=inode

End

TlimitltT(Ki)

Yes

No

No

Yes




119901119894

=




(15)

119879119897119894119898119894119905 = 1205911198640 (16)













100

90

80

70

60

50

40

30

20

10

0100908070605040302010













100

90

80

70

60

50

40

30

20

10

0

Num

ber o

f aliv

e nod

es

0

500

1000

1500

2000

2500

3000

3500

4000

4500

5000

number of rounds


(a)

90

75

60

45

30

15

Num

ber o

f pac

kets

rece

ived

by

BS

105

0

500

1000

1500

2000

2500

3000

3500

4000

4500

5000

number of rounds

EDEECDEEC

EDDEECProposed

(b)

900

800

700

600

500

400

300

200

100

0

thro

ughp

ut in

A b

ites

1000

0

500

1000

1500

2000

2500

3000

3500

4000

4500

5000

number of rounds

EDEECDEEC

EDDEECProposed

(c)

20

18

16

14

12

10

8

6

4

2

0

resid

ual e

nerg

y of

the n

etw

ork

(J)

0

500

1000

1500

2000

2500

3000

3500

4000

4500

5000

number of rounds

EDEECDEEC

EDDEECProposed

(d)1000

900

800

700

600

500

400

300

200

100

aver

age l

aten

cy (m

s)

500

1000

1500

2000

2500

3000

3500

4000

4500

5000

number of rounds

EDEECDEEC

EDDEECProposed

(e)







100

90

80

70

60

50

40

30

20

10

0100908070605040302010












5 Conclusions



100

90

80

70

60

50

40

30

20

10

0

Num

ber o

f aliv

e nod

es

0

500

1000

1500

2000

2500

3000

3500

4000

4500

5000

number of rounds


(a)

75

60

45

30

15

0

Num

ber o

f pac

kets

rece

ived

by

BS

105

0

500

1000

1500

2000

2500

3000

3500

4000

4500

5000

number of rounds

EDEECDEEC

EDDEECProposed

105

90

(b)

900

800

700

600

500

400

300

200

100

0

thro

ughp

ut in

A b

ites

1000

0

500

1000

1500

2000

2500

3000

3500

4000

4500

5000

number of rounds

EDEECDEEC

EDDEECProposed

(c)

20

18

16

14

12

10

8

6

4

2

0

resid

ual e

nerg

y of

the n

etw

ork

(J)

0

500

1000

1500

2000

2500

3000

3500

4000

4500

5000

number of rounds

EDEECDEEC

EDDEECProposed

(d)

1000

900

800

700

600

500

400

300

200

100

aver

age l

aten

cy (m

s)

500

1000

1500

2000

2500

3000

3500

4000

4500

5000

number of rounds

EDEECDEEC

EDDEECProposed

(e)







100

90

80

70

60

50

40

30

20

10

0

num

ber o

f aliv

e nod

es

0

500

1000

1500

2000

2500

3000

3500

4000

4500

5000

number of rounds(a)

20

18

16

14

12

10

8

6

4

2

0

resid

ual e

nerg

y of

the n

etw

ork

(J)

0

500

1000

1500

2000

2500

3000

3500

4000

4500

5000

number of rounds(b)



Data Availability




References
































RoboticsJournal of




VLSI Design



Shock and Vibration







Journal of



Volume 2018



Volume 2018


Journal of




SensorsJournal of



RotatingMachinery





Propagation






Hindawi


Advances in

Multimedia



Start

Ei(r)gt Tlimit



CH=inode

End

TlimitltT(Ki)

Yes

No

No

Yes




119901119894

=




(15)

119879119897119894119898119894119905 = 1205911198640 (16)













100

90

80

70

60

50

40

30

20

10

0100908070605040302010













100

90

80

70

60

50

40

30

20

10

0

Num

ber o

f aliv

e nod

es

0

500

1000

1500

2000

2500

3000

3500

4000

4500

5000

number of rounds


(a)

90

75

60

45

30

15

Num

ber o

f pac

kets

rece

ived

by

BS

105

0

500

1000

1500

2000

2500

3000

3500

4000

4500

5000

number of rounds

EDEECDEEC

EDDEECProposed

(b)

900

800

700

600

500

400

300

200

100

0

thro

ughp

ut in

A b

ites

1000

0

500

1000

1500

2000

2500

3000

3500

4000

4500

5000

number of rounds

EDEECDEEC

EDDEECProposed

(c)

20

18

16

14

12

10

8

6

4

2

0

resid

ual e

nerg

y of

the n

etw

ork

(J)

0

500

1000

1500

2000

2500

3000

3500

4000

4500

5000

number of rounds

EDEECDEEC

EDDEECProposed

(d)1000

900

800

700

600

500

400

300

200

100

aver

age l

aten

cy (m

s)

500

1000

1500

2000

2500

3000

3500

4000

4500

5000

number of rounds

EDEECDEEC

EDDEECProposed

(e)







100

90

80

70

60

50

40

30

20

10

0100908070605040302010












5 Conclusions



100

90

80

70

60

50

40

30

20

10

0

Num

ber o

f aliv

e nod

es

0

500

1000

1500

2000

2500

3000

3500

4000

4500

5000

number of rounds


(a)

75

60

45

30

15

0

Num

ber o

f pac

kets

rece

ived

by

BS

105

0

500

1000

1500

2000

2500

3000

3500

4000

4500

5000

number of rounds

EDEECDEEC

EDDEECProposed

105

90

(b)

900

800

700

600

500

400

300

200

100

0

thro

ughp

ut in

A b

ites

1000

0

500

1000

1500

2000

2500

3000

3500

4000

4500

5000

number of rounds

EDEECDEEC

EDDEECProposed

(c)

20

18

16

14

12

10

8

6

4

2

0

resid

ual e

nerg

y of

the n

etw

ork

(J)

0

500

1000

1500

2000

2500

3000

3500

4000

4500

5000

number of rounds

EDEECDEEC

EDDEECProposed

(d)

1000

900

800

700

600

500

400

300

200

100

aver

age l

aten

cy (m

s)

500

1000

1500

2000

2500

3000

3500

4000

4500

5000

number of rounds

EDEECDEEC

EDDEECProposed

(e)







100

90

80

70

60

50

40

30

20

10

0

num

ber o

f aliv

e nod

es

0

500

1000

1500

2000

2500

3000

3500

4000

4500

5000

number of rounds(a)

20

18

16

14

12

10

8

6

4

2

0

resid

ual e

nerg

y of

the n

etw

ork

(J)

0

500

1000

1500

2000

2500

3000

3500

4000

4500

5000

number of rounds(b)



Data Availability




References
































RoboticsJournal of




VLSI Design



Shock and Vibration







Journal of



Volume 2018



Volume 2018


Journal of




SensorsJournal of



RotatingMachinery





Propagation






Hindawi


Advances in

Multimedia





100

90

80

70

60

50

40

30

20

10

0100908070605040302010













100

90

80

70

60

50

40

30

20

10

0

Num

ber o

f aliv

e nod

es

0

500

1000

1500

2000

2500

3000

3500

4000

4500

5000

number of rounds


(a)

90

75

60

45

30

15

Num

ber o

f pac

kets

rece

ived

by

BS

105

0

500

1000

1500

2000

2500

3000

3500

4000

4500

5000

number of rounds

EDEECDEEC

EDDEECProposed

(b)

900

800

700

600

500

400

300

200

100

0

thro

ughp

ut in

A b

ites

1000

0

500

1000

1500

2000

2500

3000

3500

4000

4500

5000

number of rounds

EDEECDEEC

EDDEECProposed

(c)

20

18

16

14

12

10

8

6

4

2

0

resid

ual e

nerg

y of

the n

etw

ork

(J)

0

500

1000

1500

2000

2500

3000

3500

4000

4500

5000

number of rounds

EDEECDEEC

EDDEECProposed

(d)1000

900

800

700

600

500

400

300

200

100

aver

age l

aten

cy (m

s)

500

1000

1500

2000

2500

3000

3500

4000

4500

5000

number of rounds

EDEECDEEC

EDDEECProposed

(e)







100

90

80

70

60

50

40

30

20

10

0100908070605040302010












5 Conclusions



100

90

80

70

60

50

40

30

20

10

0

Num

ber o

f aliv

e nod

es

0

500

1000

1500

2000

2500

3000

3500

4000

4500

5000

number of rounds


(a)

75

60

45

30

15

0

Num

ber o

f pac

kets

rece

ived

by

BS

105

0

500

1000

1500

2000

2500

3000

3500

4000

4500

5000

number of rounds

EDEECDEEC

EDDEECProposed

105

90

(b)

900

800

700

600

500

400

300

200

100

0

thro

ughp

ut in

A b

ites

1000

0

500

1000

1500

2000

2500

3000

3500

4000

4500

5000

number of rounds

EDEECDEEC

EDDEECProposed

(c)

20

18

16

14

12

10

8

6

4

2

0

resid

ual e

nerg

y of

the n

etw

ork

(J)

0

500

1000

1500

2000

2500

3000

3500

4000

4500

5000

number of rounds

EDEECDEEC

EDDEECProposed

(d)

1000

900

800

700

600

500

400

300

200

100

aver

age l

aten

cy (m

s)

500

1000

1500

2000

2500

3000

3500

4000

4500

5000

number of rounds

EDEECDEEC

EDDEECProposed

(e)







100

90

80

70

60

50

40

30

20

10

0

num

ber o

f aliv

e nod

es

0

500

1000

1500

2000

2500

3000

3500

4000

4500

5000

number of rounds(a)

20

18

16

14

12

10

8

6

4

2

0

resid

ual e

nerg

y of

the n

etw

ork

(J)

0

500

1000

1500

2000

2500

3000

3500

4000

4500

5000

number of rounds(b)



Data Availability




References
































RoboticsJournal of




VLSI Design



Shock and Vibration







Journal of



Volume 2018



Volume 2018


Journal of




SensorsJournal of



RotatingMachinery





Propagation






Hindawi


Advances in

Multimedia



100

90

80

70

60

50

40

30

20

10

0

Num

ber o

f aliv

e nod

es

0

500

1000

1500

2000

2500

3000

3500

4000

4500

5000

number of rounds


(a)

90

75

60

45

30

15

Num

ber o

f pac

kets

rece

ived

by

BS

105

0

500

1000

1500

2000

2500

3000

3500

4000

4500

5000

number of rounds

EDEECDEEC

EDDEECProposed

(b)

900

800

700

600

500

400

300

200

100

0

thro

ughp

ut in

A b

ites

1000

0

500

1000

1500

2000

2500

3000

3500

4000

4500

5000

number of rounds

EDEECDEEC

EDDEECProposed

(c)

20

18

16

14

12

10

8

6

4

2

0

resid

ual e

nerg

y of

the n

etw

ork

(J)

0

500

1000

1500

2000

2500

3000

3500

4000

4500

5000

number of rounds

EDEECDEEC

EDDEECProposed

(d)1000

900

800

700

600

500

400

300

200

100

aver

age l

aten

cy (m

s)

500

1000

1500

2000

2500

3000

3500

4000

4500

5000

number of rounds

EDEECDEEC

EDDEECProposed

(e)







100

90

80

70

60

50

40

30

20

10

0100908070605040302010












5 Conclusions



100

90

80

70

60

50

40

30

20

10

0

Num

ber o

f aliv

e nod

es

0

500

1000

1500

2000

2500

3000

3500

4000

4500

5000

number of rounds


(a)

75

60

45

30

15

0

Num

ber o

f pac

kets

rece

ived

by

BS

105

0

500

1000

1500

2000

2500

3000

3500

4000

4500

5000

number of rounds

EDEECDEEC

EDDEECProposed

105

90

(b)

900

800

700

600

500

400

300

200

100

0

thro

ughp

ut in

A b

ites

1000

0

500

1000

1500

2000

2500

3000

3500

4000

4500

5000

number of rounds

EDEECDEEC

EDDEECProposed

(c)

20

18

16

14

12

10

8

6

4

2

0

resid

ual e

nerg

y of

the n

etw

ork

(J)

0

500

1000

1500

2000

2500

3000

3500

4000

4500

5000

number of rounds

EDEECDEEC

EDDEECProposed

(d)

1000

900

800

700

600

500

400

300

200

100

aver

age l

aten

cy (m

s)

500

1000

1500

2000

2500

3000

3500

4000

4500

5000

number of rounds

EDEECDEEC

EDDEECProposed

(e)







100

90

80

70

60

50

40

30

20

10

0

num

ber o

f aliv

e nod

es

0

500

1000

1500

2000

2500

3000

3500

4000

4500

5000

number of rounds(a)

20

18

16

14

12

10

8

6

4

2

0

resid

ual e

nerg

y of

the n

etw

ork

(J)

0

500

1000

1500

2000

2500

3000

3500

4000

4500

5000

number of rounds(b)



Data Availability




References
































RoboticsJournal of




VLSI Design



Shock and Vibration







Journal of



Volume 2018



Volume 2018


Journal of




SensorsJournal of



RotatingMachinery





Propagation






Hindawi


Advances in

Multimedia





100

90

80

70

60

50

40

30

20

10

0100908070605040302010












5 Conclusions



100

90

80

70

60

50

40

30

20

10

0

Num

ber o

f aliv

e nod

es

0

500

1000

1500

2000

2500

3000

3500

4000

4500

5000

number of rounds


(a)

75

60

45

30

15

0

Num

ber o

f pac

kets

rece

ived

by

BS

105

0

500

1000

1500

2000

2500

3000

3500

4000

4500

5000

number of rounds

EDEECDEEC

EDDEECProposed

105

90

(b)

900

800

700

600

500

400

300

200

100

0

thro

ughp

ut in

A b

ites

1000

0

500

1000

1500

2000

2500

3000

3500

4000

4500

5000

number of rounds

EDEECDEEC

EDDEECProposed

(c)

20

18

16

14

12

10

8

6

4

2

0

resid

ual e

nerg

y of

the n

etw

ork

(J)

0

500

1000

1500

2000

2500

3000

3500

4000

4500

5000

number of rounds

EDEECDEEC

EDDEECProposed

(d)

1000

900

800

700

600

500

400

300

200

100

aver

age l

aten

cy (m

s)

500

1000

1500

2000

2500

3000

3500

4000

4500

5000

number of rounds

EDEECDEEC

EDDEECProposed

(e)







100

90

80

70

60

50

40

30

20

10

0

num

ber o

f aliv

e nod

es

0

500

1000

1500

2000

2500

3000

3500

4000

4500

5000

number of rounds(a)

20

18

16

14

12

10

8

6

4

2

0

resid

ual e

nerg

y of

the n

etw

ork

(J)

0

500

1000

1500

2000

2500

3000

3500

4000

4500

5000

number of rounds(b)



Data Availability




References
































RoboticsJournal of




VLSI Design



Shock and Vibration







Journal of



Volume 2018



Volume 2018


Journal of




SensorsJournal of



RotatingMachinery





Propagation






Hindawi


Advances in

Multimedia



100

90

80

70

60

50

40

30

20

10

0

Num

ber o

f aliv

e nod

es

0

500

1000

1500

2000

2500

3000

3500

4000

4500

5000

number of rounds


(a)

75

60

45

30

15

0

Num

ber o

f pac

kets

rece

ived

by

BS

105

0

500

1000

1500

2000

2500

3000

3500

4000

4500

5000

number of rounds

EDEECDEEC

EDDEECProposed

105

90

(b)

900

800

700

600

500

400

300

200

100

0

thro

ughp

ut in

A b

ites

1000

0

500

1000

1500

2000

2500

3000

3500

4000

4500

5000

number of rounds

EDEECDEEC

EDDEECProposed

(c)

20

18

16

14

12

10

8

6

4

2

0

resid

ual e

nerg

y of

the n

etw

ork

(J)

0

500

1000

1500

2000

2500

3000

3500

4000

4500

5000

number of rounds

EDEECDEEC

EDDEECProposed

(d)

1000

900

800

700

600

500

400

300

200

100

aver

age l

aten

cy (m

s)

500

1000

1500

2000

2500

3000

3500

4000

4500

5000

number of rounds

EDEECDEEC

EDDEECProposed

(e)







100

90

80

70

60

50

40

30

20

10

0

num

ber o

f aliv

e nod

es

0

500

1000

1500

2000

2500

3000

3500

4000

4500

5000

number of rounds(a)

20

18

16

14

12

10

8

6

4

2

0

resid

ual e

nerg

y of

the n

etw

ork

(J)

0

500

1000

1500

2000

2500

3000

3500

4000

4500

5000

number of rounds(b)



Data Availability




References
































RoboticsJournal of




VLSI Design



Shock and Vibration







Journal of



Volume 2018



Volume 2018


Journal of




SensorsJournal of



RotatingMachinery





Propagation






Hindawi


Advances in

Multimedia





100

90

80

70

60

50

40

30

20

10

0

num

ber o

f aliv

e nod

es

0

500

1000

1500

2000

2500

3000

3500

4000

4500

5000

number of rounds(a)

20

18

16

14

12

10

8

6

4

2

0

resid

ual e

nerg

y of

the n

etw

ork

(J)

0

500

1000

1500

2000

2500

3000

3500

4000

4500

5000

number of rounds(b)



Data Availability




References
































RoboticsJournal of




VLSI Design



Shock and Vibration







Journal of



Volume 2018



Volume 2018


Journal of




SensorsJournal of



RotatingMachinery





Propagation






Hindawi


Advances in

Multimedia





















RoboticsJournal of




VLSI Design



Shock and Vibration







Journal of



Volume 2018



Volume 2018


Journal of




SensorsJournal of



RotatingMachinery





Propagation






Hindawi


Advances in

Multimedia


threshold balanced sampled deec model for heterogeneous...

Documents