an enhanced pegasis algorithm with mobile sink support for...

Research ArticleAn Enhanced PEGASIS Algorithm with MobileSink Support for Wireless Sensor Networks

JinWang 1 Yu Gao2 Xiang Yin2 Feng Li1 and Hye-Jin Kim 3

1School of Computer amp Communication Engineering Changsha University of Science amp Technology China2School of Information Engineering Yangzhou University China3Business Administration Research Institute Sungshin W University Republic of Korea

Correspondence should be addressed to Hye-Jin Kim hye-jinkimhotmailcom

Received 4 May 2018 Revised 1 November 2018 Accepted 18 November 2018 Published 3 December 2018

Academic Editor Jaime Lloret

Copyright copy 2018 JinWang et alThis is an open access article distributed under the Creative Commons Attribution License whichpermits unrestricted use distribution and reproduction in any medium provided the original work is properly cited

Energy efficiency has been a hot research topic for many years andmany routing algorithms have been proposed to improve energyefficiency and to prolong lifetime for wireless sensor networks (WSNs) Since nodes close to the sink usually need to consumemoreenergy to forward data of its neighbours to sink they will exhaust energy more quickly These nodes are called hot spot nodes andwe call this phenomenon hot spot problem In this paper an Enhanced Power Efficient Gathering in Sensor Information Systems(EPEGASIS) algorithm is proposed to alleviate the hot spots problem from four aspects Firstly optimal communication distanceis determined to reduce the energy consumption during transmission Then threshold value is set to protect the dying nodes andmobile sink technology is used to balance the energy consumption among nodes Next the node can adjust its communicationrange according to its distance to the sink node Finally extensive experiments have been performed to show that our proposedEPEGASIS performs better in terms of lifetime energy consumption and network latency

1 Introduction

Topology control is a very important research issue for theemerging mobile networks (EMN)Wireless sensor networks(WSNs) as part of the EMN are usually composed of alarge collection of tiny sensors nodes and they are developingvery rapidly in recently years due to their wide applications[1] These nodes are usually deployed in a random way andthey can collect information from surroundings and thentransfer the data package to sink node using single ormultiplehops communication to formWSNsThe outstanding perfor-mance of WSNs like fault tolerance rapid deployment self-organizing timely response etc makes WSNs widely used inharsh environment such as military surveillance industrialproduct line monitoring medical health care and smarthomes [2 3]

Energy efficiency and energy balancing have been a veryhot and challenging research issue for WSNs for many yearsSince large number of sensor nodes is deployed under veryharsh environment it is unrealistic to change the sensor

batteries for them Besides the nodes close to the sinknot only need to collect data but also need to forward itsneighboursrsquo data Thus those nodes will exhaust their energymore quickly than other nodes far away from sink Thisphenomenon is known as hot spots problem When the firstnode dies because of energy exhausting the performance ofnetwork in terms of connectivity coverage lifetime etc willdecrease sharply [4 5]

In order to solve the hot spots problem sink mobilitytechnology is introduced to many routing protocols [6ndash9]By adopting sink mobility the following advantages can beachieved Firstly the mobile sink moves along a certaintrajectory so that the nodes close to the sink could taketurns to be the forwarder which will largely alleviate hotspots problem [10] Secondly the energy consumption of thewhole network can be reduced because of the shorter averagetransmission distance between sensor nodes andmobile sinkif the mobile sink has a proper trajectory [11] Thirdly theperformance of the network in terms of transmission latencyand network throughput can be greatly improved under

HindawiWireless Communications and Mobile ComputingVolume 2018 Article ID 9472075 9 pageshttpsdoiorg10115520189472075

2 Wireless Communications and Mobile Computing

suitable moving strategy Finally mobile sink can enhance thenetwork connectivity in sparse network [12ndash14]

The main contribution in this paper includes the follow-ing four aspects First we find the optimal communicationdistance for data transmission Second we set threshold valuefor each node to protect those nodes with low remainingenergy Third we adjust nodesrsquo communication distanceaccording to its distance to the mobile sink Finally mobilesink technology is adopted to balance the energy of differentregions Numerous simulations are conducted to prove ourproposed algorithm outperforms than some existing work

The rest of the paper is organized as follows In Section 1we introduce the background of our work In Section 2we discuss some classic routing protocols and some latestresearch achievement The system model which containsnetwork and energy model is presented in Section 3 InSection 4 we describe our algorithm in detail Extensivesimulations are conducted and the experimental results arediscussed and compared in Section 5 Section 6 concludes thepaper with some future work

2 Related Work

Many researchers have paid close attention to energy efficientrouting protocols or algorithms in recent years LEACH(Low-Energy Adaptive Clustering Algorithm) is one of themost famous hierarchical protocols which were proposedabout twenty years ago [15] In LEACH there are two typesof nodes namely cluster heads (CHs) and ordinary nodes(ONs) Each ON collects data from area of interest and sendsthe data package to a closest CH Each CH take charge offusing the data it receives and then transmits the fused datato the sink node The introduction of CHs can avoid longdistance communication between ON and sink node thusmuch energy is saved However the selection of CHs is ina random way as a result of random selection So the CHdistributes unevenly inside the whole network with degradedperformance Meanwhile CHs communicate with sink nodedirectly which causes much energy dissipation

PEGASIS (Power Efficient Gathering in Sensor Informa-tion Systems) is a chain-based routing protocol for WSNs[16] In PEGASIS each sensor node only needs to transmitdata to its neighbour which is closer to the sink nodeSeveral chains could be constructed according to the greedyalgorithmand the leader of each chain takes the responsibilityto transfer the data to the sink node Due to the heavy burdenof the leaders of chains each node takes turns to be the leaderto balance the energy consumption Because of multihoppropagation long distance communication between sensornode and sink node is avoided and much energy is savedMeanwhile due to multihop propagation the delay of thenetwork is very serious and it is not suitable for delay sensitiveapplications

In [17] the authors propose an energy efficient routingprotocol using mobile sink based on clustering and it issuitable forWSNswith obstruction Mobile sink moves alongthe CHs and collects data by single hop communicationIn this protocol an efficient scheduling mechanism basedon spanning graphs is proposed to find a shortest path for

mobile sink to avoid obstacle Simulation result shows thatthe lifetime of the network is prolonged and complexity ofnetwork is reduced

In [18] the authors proposed a data collection algorithmusing mobile sink based on tree clustering This algorithmcontains three phases namely tree construction tree decom-position and subrendezvous points (SRP) selection and datacollection phases Mobile sink moves towards rendezvouspoints (RP) and SRP to collect data via single hop commu-nication After data collection in each round sensor nodeswill reselect RP and SRP Simulation result proves that theproposed algorithm performs better in terms of networklifetime and the moving path of the mobile sink is short

In [19] the authors study the event-driven applicationfor WSNs with mobile sinks In this paper each sensornode has two statuses monitoring and transmission statusWhen an event is caught the status of node changes frommonitoring to transmission and the group of active sensornodes (ASN) is constructed to forward the data to the mobilesink In large-scale WSNs the area of interest is divided intoseveral subareas due to the limit of the speed of the mobilesink Each subarea contains a mobile sink and ASNs areselected to collect and transmit data In the meantime thecontinuous and optimal trajectory (COT) can be calculatedfor the mobile sink to achieve better performance

In [20] the authors propose an energy efficient routingprotocol formobile sensor networks using a path-constrainedmobile sink This protocol is suitable for WSNs which limitson the moving path of the mobile sink such as mobile sinksdeployed in bus Source node sends the data package to thetarget node which is closest to the location where the mobilesink will arrive next time and ensure the shortest path totransmit data This kind of protocol is very suitable for delay-tolerant network and it possesses higher robustness and lowerenergy consumption

In addition some heuristic algorithms are used toimprove the performance of the network [21ndash26] In [21]the authors combine the Particle Swarm Optimization (PSO)algorithms with cluster technology to enhance the net-work lifetime In [22] the authors propose an Ant ColonyOptimization (ACO) based on clustering algorithm to findan optimal moving trajectory for mobile sink In [23]techniques like Glow-worm Swarm Optimization (GSO)clustering and mobile sink are combined together to improveenergy efficiency as well as to prolong the lifetime of wirelesssensor network

3 System Model

31 Basic Assumptions In this paper we make some basicassumptions as follows

(1) All the sensor nodes are randomly deployed and keepstatic after deployment

(2) Each sensor node has a unique ID to differ from eachother

(3) All the sensor nodes have the same initial energy andbatteries of them cannot be changed

Wireless Communications and Mobile Computing 3

mobile sinksensor node

Figure 1 Network model

(4) Mobile sink can move freely and owns unconstrainedenergy

(5) Transmission power of sensors can be adjusted basedon communication distance [27]

32 Network Model In this paper N nodes are deployedrandomly in a circular field with radius R These sensornodes are denoted as 1198991 1198992 1198993 sdot sdot sdot 119899119899 respectively as isshown in Figure 1 In each round every sensor node shouldtransmit one package to the sink using single hop ormultihopcommunication based on the relatively distance

33 Energy Model The first ratio model is adopted here tocalculate the energy consumption [28 29] As is shown inFigure 2 the energy consumption mainly contains two partswhich are transmission consumption and reception energyconsumption

Transmission consumption mainly includes the energyconsumed for transmission circuit and the power amplifierexpenditureThe reception consumption mainly includes theenergy of reception circuit expenditure Transmission con-sumption can be calculated by using the following formula

119864119879119909 (119897 119889) = 119897 sdot 119864119890119897119890119888 + 119897 sdot 120576119891119904 sdot 1198892 if 119889 lt 1198890119897 sdot 119864119890119897119890119888 + 119897120576119898119901 sdot 1198894 if 119889 ge 1198890

(1)

where 119864119890119897119890119888 is the energy cost to use transmitter or receivercircuit to process one-bit signal Symbols 120576119891119904 and 120576119898119901 denotethe amplification coefficient for the free space model and themultipath fading model separately Metric 1198890 is the thresholdvalue and can be calculated as 1198890 = radic120576119891119904120576119898119901

The energy consumption for reception can be calculatedaccording to formula (2)

119864119877119909 (119897) = 119897 sdot 119864119890119897119890119888 (2)

4 Our Proposed Algorithm

41 Overview of EPEGASIS In this section we will illustrateour proposed algorithm in detail Our proposed algorithmis called Enhanced PEGASIS (EPEGASIS) which belongsto the chain-based routing algorithms In EPEGASIS eachnode uses optimal communication distance to choose therelay node from its neighbours for data transmission In orderto avoid excessive energy consumption for some nodes withspecial location we set a protection mechanism accordingto the average residual energy of its neighbours A mobilesink is utilized to gather data for different region energyconsumption balancing The workflow of each node in thenetwork is shown as Figure 3

42 Study on Optimal Communication Distance Accordingto the energy consumption model the energy consumptionraises rapidly with the increasing of the communicationdistance In order to conserve energy multihop transmissionis adopted to avoid long distance communication Howevertoo much hops may increase the burden of forwarding andcause increased end-to-end delay or latency

We assume that the source node is 119898 meters away fromthe mobile sink and it sends a one-bit data package to themobile sink by 119896 hops and the total energy consumption canbe calculated using formula (3)

119864119905119900119905119886119897

=

119896[119864119890119897119890119888 + 119864119891119904 (119898119896 )2] + (119896 minus 1) 119864119890119897119890119888 if 119898

119896 lt 1198890119896 [119864119890119897119890119888 + 119864119898119901 (119898

119896 )4] + (119896 minus 1) 119864119890119897119890119888 if 119898119896 ge 1198890

(3)

The relationship between total energy consumption andhop counts can be shown as Figure 4 From Figure 4 wecan clearly see that as the hop counts increases the totalenergy consumption drops first to reach its lowest pointand then rises The triangle in each line denotes the energyconsumption when 119898119896 = 1198890 and it also denotes the lowestpoint of each line Therefore the optimal communicationdistance and optimal hop counts is shown as formulas (4) and(5)

119889119900119901119905119894119898119886119897 = 1198890 (4)

119900119901119905119894119898119886119897 ℎ119900119901 119888119900119906119899119905119904 = lceil1198981198890rceil (5)

43 Initial Phase of the Network After nodes deployment thenetwork enters into initial phase During the initial phasethe main object is to exchange information to prepare fordata transmission At the very beginning the mobile sinkbroadcast a NETWORK INIT package to remind all thenodes to initial the network Each node maintains a routetable called ldquoNBR TABLErdquo to record its neighboursrsquo infor-mation Then each node broadcast a NODE INFO packagewith its maximal transmission distance Neighbours whoreceive NODE INFO package will write relevant informationinto route table After the initial phase each node will have


Table 1 Package information

Package type Content

NETWORK INIT The initial location of the mobile sink the moving speed and direction of the mobile sinktime of duration of each round TDMA schedule of nodes

NODE INFO The location the residual energy ID of source nodeENERGY INFO The residual energy and ID of source nodeSINK INFO System time the current location of mobile sink the speed and direction of mobile sink

Figure 2 Energy model

the knowledge of its neighbours In order to update theenergy information of sensors an ENERGY INFO packageis broadcasted by each node during each round With therunning of the network the system time and the locationof the mobile sink may go awry so the mobile sink willbroadcast a SINK INFO package in the whole networkat fixed periods The detailed information of packages isillustrated in Table 1

44 Network Topology Generation In PEGASIS each nodeselects its closest neighbour as relay node for data forwardingand that causes heavy burden of transmission and long net-work delay In our proposed algorithm we mainly decreasethe times of data forwarding and achieve an optimal energyconsumingWe use optimal communication and optimal hopcounts for data transmissionNode selects a forwarder amongits routing table and the forwarder satisfies the following twoformulas

119889 (119899119891119900119903119908119886119903119889119890119903mobile sink) lt 119889 (119899119904119900119906119903119888119890mobile sink) (6)

119889 (119899119904119900119906119903119888119890 119899119891119900119903119908119886119903119889119890119903)= min 10038161003816100381610038161003816119889 (119899119904119900119906119903119888119890 119899119891119900119903119908119886119903119889119890119903) minus 11988911990011990111990511989411989811988611989710038161003816100381610038161003816

(7)

where 119889(119899119894 119899119895) denotes the distance between node 119894 and node119895When the mobile sink is in the optimal communication

of sensors they will transmit data to the mobile sink directlyAfter relay node selection nodes far away from the mobilesink transmit data to it by several relays and the topologyof the network is generated The topology of the network isshown as Figure 5

45 Protection Mechanism for Dying Node Nodes close tothe sensor field centre take on a heavy burden for dataforwarding however nodes close to the edge of the sensorfield only need to send their own data and that cause the hot

spots problem Figure 6 describes the hot spots phenomenonafter the network running for several rounds

In order to protect nodes from dying too early we seta threshold value for nodes according to average neighbourenergy When nodersquos residual energy is less than a thresholdvalue it will not take the role of forwarder and only sends itsown generated data The threshold value is calculated usingformula (8)

119864119905ℎ119903119890119904ℎ119900119897119889 = sum119899isin119862119894 119864119903119890119904119894119889119906119886119897119873 (8)

where 119862119894 denotes the set of the neighbours of node 119894119873 is thenumber of its neighbours and 119864119903119890119904119894119889119906119886119897 is the current energyof node 119899

By means of setting threshold value nodes in the centralarea are protected and time of first node die is greatly delayedDuring the initial round all nodesrsquo residual energy reachthreshold value and all of them can be chosen as relay nodeWith round increasing nodes close to the central area ownlower residual energy than threshold value therefore nodesclose to the edge will gradually become relay nodes and causering routing The ring routing is illustrated as Figure 7

46 Communication Distance Adjustment for Edge NodeNodes in the edge area hardly need to be the forwarderand it causes the unbalanced energy consumption amongdifferent region In order to take full advantage of edgenodes communication distance adjustment is introducedto further balance the energy consumption Namely thecommunication distance of nodes is adjusted according tothe distance to the mobile sink Nodes far away from themobile sink use longer communication distance for fullyenergy utilizing and nodes close to the mobile sink useshorter transmission distance to reduce energy consumption


Figure 3 Workflow of nodes

0 5 10 15Hops

0

1

2

3

4

5

Ener

gy co

nsum

ptio

n

m=300m=400m=500

times10-6

Figure 4 Energy consumption of different hop counts

The communication distance of node 119894 can be calculatedaccording to formula (9)

119889119886119889119895119906119904119905 119900119901119905119894119898119886119897 = 119889119900119901119905119894119898119886119897+ (119889 (119899119904119900119906119903119888119890mobile sink) minus 119877

2 ) sdot 120572 (9)

where 119877 is the radius of the network and 120572 (as is discussedin next section) is the adjusting parameter The chain struc-ture after communication distance adjustment is shown asFigure 8

Figure 5 Network topology

47 Mobile Sink Moving Schema In our proposed algorithmthe mobile sink moves with a predetermined direction andspeed The mobile sink moves around the centre of the circlewith a constant radius (as is discussed in next section)Therefore themobile sink only needs to broadcast its positionin initial phase After119905 time interval the mobile sinkmovesto a new position as is shown in Figure 9

5 Performance Evaluation

51 Simulation Environment In order to evaluate the perfor-mance of our proposed EPEGASIS algorithm we use Matlab



Figure 7 Ring routing

Figure 8 Chain structure after communication distance adjust-ment

Figure 9 Moving schema of mobile sink

Comparison of Nodes Alive

PEGASISEPEGASIS1EPEGASIS2

50 100 150 2000Rounds

50

100

150

200

250

300

350

400

450

500N

odes

Aliv

e

Figure 10 Network lifetime comparison under different routingprotocols

simulator to conduct the experiment We will compareour proposed algorithm with typical PEGASIS And thesimulation parameters are listed in Table 2

52 Network Lifetime Analysis As is mentioned abovePEGASIS is a classic chain-based routing protocol and wewill compare our algorithm with itThe protocol we proposedwithout communication range adaption is called EPEGASIS1and the protocol with further improvement which introducescommunication distance adjustment is called EPEGASIS2The simulation results are shown in Figure 10


Comparison of average energy consumption(J)


20 30 40 50 60 70 80 90 10010Rounds

0

1

2

3

4

5

Aver

age e

nerg

y co

nsum

ptio

n

times10-4

Figure 11 Average energy consumption comparison under differentrouting protocols

Table 2 Simulation parameters

Parameter Name Parameter ValueNetwork radius(R) 300mMobile sink radius(r) [0 75 150 225 300]mMobile sink speed(w) pi5Number of nodes (N) 500Packet length (l) 500 bitsInitial energy (1198640) 005 JEnergy consumption on circuit(119864119890119897119890119888) 50nJbit

Free-space channel parameter (120576119891119904) 10pJbitm2Multi-path channel parameter (120576119898119901) 00013pJbitm4Adjusting parameter (120572) [01 02 03 04]

Figure 10 describes that the lifetime of the network isgreatly improved in EPEGASIS1 compared to PEGASIS Dueto the heavy burden to transmit the data packages of neigh-bours nodes close to the sink begin to die at about 60 roundsHowever the first dies are at about 90 rounds in EPEGASIS2because of the communication distance adjustment nodesprotection and sink mobility

53 Energy Consumption Analysis We compare the averageenergy consumption per round of different routing protocolsand the result is shown as Figure 11We can obviously see thatthe average energy consumption per round in EPEGASIS1reduces about one-third that of PEGASIS due to the using ofoptimal communication distance

Comparison of average hops


0

2

4

6

8

10

12

14

16

18

20

Aver

age h

ops

10 20 30 40 50 60 70 80 90 1000Rounds

Figure 12 Average hops under different routing protocols

54 Study on Network Latency We also study the networklatency between different routing protocols and the resultis shown as Figure 12 We can clearly see that the networklatency has a great improvement in both EPEGASIS1 andEPEGASIS2 In EPEGASIS2 due to the adoption of com-munication distance adjustment some nodes extend theircommunication distance and some nodes short their com-munication distanceTherefore EPEGASIS1 and EPEGASIS2almost have the same performance in terms of networklatency

55 Study on Communication Range Adaption As we dis-cussed above each sensor node adjusts its communicationdistance in accordance with the distance to the mobilesink and the adjusting parameter 120572 determines the detailedchanges From Figure 13 we can obviously see that when120572=02 the network has a better performance in terms ofnetwork lifetime

56 Study on Mobile Sink Moving Trajectory The movingtrajectory of the mobile sink has a significant influence onthe lifetime of the network In our proposed PEGASIS2the mobile sink moves around the centre of the circle andwe change the radius of the mobile sink to enhance theperformance of the network Simulation result is shown asFigure 14 and it demonstrates that when the mobile sinktravels around the sensor filed with 025R radius the networkhas a better performance in terms of network lifetime

6 Conclusions

In this paper we proposed an Enhanced PEGASIS algorithmwith mobile sink support to improve the performance of



=01=02

=03=04

0

50

100

150

200

250

300

350

400

450

500

Nod

es A

live

50 100 150 200 250 3000Rounds

Figure 13 Network lifetime under different adjusting parameter


0025R05R

075RR

50 100 150 200 250 3000Rounds

0

50

100

150

200

250

300

350

400

450

500

Nod

es A

live

Figure 14 Network lifetime under different mobile sink movingtrajectory

WSNs Nodes use optimal communication distance to selectthe forwarder and threshold value is set to protect nodes fromdying too early In order to further improve the performanceof the network we adjust the communication distance ofnodes based on their distance to the mobile sink Thereforenodes close to themobile sinkwill have short communicationdistance to decrease its energy consumption and the time offirst node dies prolongs greatly Extensive simulation resultsvalidate the performance of our proposed algorithms than

traditional PEGASIS However the hot spots problem isnot completely eliminated Our future work mainly includesproper design of sink moving trajectory as well as jointoptimization of both routing algorithm and sink trajectorydesign

Data Availability

The data used to support the findings of this study areavailable from the corresponding author upon request

Disclosure

This paper is a revised and expanded version of a paperentitled ldquoAn Enhanced PEGASIS Algorithm with MobileSink Support for Wireless Sensor Networksrdquo presented atICCCS 2018 China June 08-10

Conflicts of Interest

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

Acknowledgments

Thiswork is supported by the National Natural Science Foun-dation of China (61772454 and 6171101570) and by theNaturalScience Foundation of Jiangsu Province (BK20150460)

References

[1] T Arampatzis J Lygeros and S Manesis ldquoA survey of appli-cations of wireless sensors and wireless sensor networksrdquoin Proceedings of the 20th IEEE International Symposium onIntelligent Control and the 13th Mediterranean Conference onControl and Automation (ISIC rsquo05 and MED rsquo05) pp 719ndash724Limassol Cyprus June 2005

[2] T Bokareva W Hu and S Kanhere Wireless Sensor Networksfor Battlefield Surveillance 2006

[3] H Alemdar and C Ersoy ldquoWireless sensor networks forhealthcare a surveyrdquo Computer Networks vol 54 no 15 pp2688ndash2710 2010

[4] J N Al-Karaki andA E Kamal ldquoRouting techniques inwirelesssensor networks a surveyrdquo IEEE Wireless CommunicationsMagazine vol 11 no 6 pp 6ndash28 2004

[5] N A Pantazis and D D Vergados ldquoA survey on power controlissues in wireless sensor networksrdquo IEEE CommunicationsSurveys amp Tutorials vol 9 no 4 pp 86ndash107 2007

[6] C Wang S Guo and Y Yang ldquoAn optimization frameworkfor mobile data collection in energy-harvesting wireless sensornetworksrdquo IEEE Transactions on Mobile Computing vol 15 no12 pp 2969ndash2986 2016

[7] C-F Cheng and C-F Yu ldquoData Gathering in Wireless SensorNetworks A Combine-TSP-Reduce Approachrdquo IEEE Transac-tions on Vehicular Technology vol 65 no 4 pp 2309ndash23242016

[8] J-Y Chang and T-H Shen ldquoAn efficient tree-based powersaving scheme for wireless sensor networks with mobile sinkrdquoIEEE Sensors Journal vol 16 no 20 pp 7545ndash7557 2016


[9] J Wang L Zuo J Shen B Li and S Lee ldquoMultiple mobilesink-based routing algorithm for data dissemination in wirelesssensor networksrdquo Concurrency and Computation Practice andExperience vol 27 no 10 pp 2656ndash2667 2015

[10] S Sasirekha and S Swamynathan ldquoCluster-chain mobile agentrouting algorithm for efficient data aggregation in wirelesssensor networkrdquo Journal of Communications and Networks vol19 no 4 pp 392ndash401 2017

[11] V Soni and D K Mallick ldquoFuzzy logic based multihoptopology control routing protocol in wireless sensor networksrdquoMicrosystem Technologies vol 7 pp 1ndash13 2018

[12] J Wang J Cao R S Sherratt and J H Park ldquoAn improvedant colony optimization-based approach with mobile sink forwireless sensor networksrdquo The Journal of Supercomputing vol1 no 8 pp 1ndash13 2017

[13] C Wu E Zapevalova Y Chen and F Li ldquoTime Optimizationof Multiple Knowledge Transfers in the Big Data EnvironmentrdquoComputers Materials amp Continua vol 54 no 3 pp 269ndash2852018

[14] J Wang C Ju H-J Kim R S Sherratt and S Lee ldquoA mobileassisted coverage hole patching scheme based on particle swarmoptimization for WSNsrdquo Cluster Computing pp 1ndash9 2017

[15] W R Heinzelman A Chandrakasan and H Balakrish-nan ldquoEnergy-Efficient Communication Protocol for WirelessMicrosensor Networksrdquo IEEEComputer Society vol 18 p 80202000

[16] S Lindsey and C S Raghavendra ldquoPEGASIS power-efficientgathering in sensor information systemsrdquo in Proceedings of theIEEE Aerospace Conference vol 3 2003

[17] G Xie and F Pan ldquoCluster-Based Routing for the Mobile Sinkin Wireless Sensor Networks with Obstaclesrdquo IEEE Access vol4 pp 2019ndash2028 2016

[18] C Zhu SWu G Han L Shu and HWu ldquoA tree-cluster-baseddata-gathering algorithm for industrial WSNs with a mobilesinkrdquo IEEE Access vol 3 no 4 pp 381ndash396 2015

[19] F Tashtarian M H Yaghmaee Moghaddam K Sohrabyand S Effati ldquoOn maximizing the lifetime of wireless sensornetworks in event-driven applications with mobile sinksrdquo IEEETransactions on Vehicular Technology vol 64 no 7 pp 3177ndash3189 2015

[20] M T Nuruzzaman and H-W Ferng ldquoA low energy con-sumption routing protocol for mobile sensor networks with apath-constrained mobile sinkrdquo in Proceedings of the 2016 IEEEInternational Conference on Communications ICC 2016 pp 1ndash6Malaysia May 2016

[21] J Wang Y Cao B Li H-J Kim and S Lee ldquoParticle swarmoptimization based clustering algorithm with mobile sink forWSNsrdquo Future Generation Computer Systems vol 76 pp 452ndash457 2017

[22] J Wang J Cao B Li S Lee and R S Sherratt ldquoBio-inspiredant colony optimization based clustering algorithmwithmobilesinks for applications in consumer home automationnetworksrdquoIEEE Transactions on Consumer Electronics vol 61 no 4 pp438ndash444 2015

[23] JWang Y-Q Cao B Li S-Y Lee and J-U Kim ldquoA glowwormswarm optimization based clustering algorithm with mobilesink support for wireless sensor networksrdquo Journal of InternetTechnology vol 16 no 5 pp 825ndash832 2015

[24] W Wang H Shi D Wu et al ldquoVD-PSO An efficient mobilesink routing algorithm in wireless sensor networksrdquo Peer-to-Peer Networking and Applications vol 10 no 3 pp 1ndash10 2016

[25] M Azharuddin and P K Jana ldquoPSO-based approach forenergy-efficient and energy-balanced routing and clustering inwireless sensor networksrdquo Soft Computing vol 21 no 22 pp6825ndash6839 2017

[26] S Arjunan and P Sujatha ldquoLifetime maximization of wirelesssensor network using fuzzy based unequal clustering and ACObased routing hybrid protocolrdquo Applied Intelligence pp 1ndash182017

[27] W Liu X Luo Y Liu J Liu M Liu and Q Yun ldquoLocalizationAlgorithm of Indoor Wi-Fi Access Points Based on SignalStrength Relative Relationship and Region Divisionrdquo Comput-ers Materials amp Continua vol 55 no 1 pp 71ndash93 2018

[28] J Wang J Cao S Ji and J H Park ldquoEnergy-efficient cluster-based dynamic routes adjustment approach for wireless sensornetworks with mobile sinksrdquo The Journal of Supercomputingvol 73 no 7 pp 3277ndash3290 2017

[29] R Meng S G Rice J Wang and X Sun ldquoA Fusion Stegano-graphic Algorithm Based on Faster R-CNNrdquo Computers Mate-rials amp Continua vol 55 no 1 pp 1ndash16 2018

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


suitable moving strategy Finally mobile sink can enhance thenetwork connectivity in sparse network [12ndash14]

The main contribution in this paper includes the follow-ing four aspects First we find the optimal communicationdistance for data transmission Second we set threshold valuefor each node to protect those nodes with low remainingenergy Third we adjust nodesrsquo communication distanceaccording to its distance to the mobile sink Finally mobilesink technology is adopted to balance the energy of differentregions Numerous simulations are conducted to prove ourproposed algorithm outperforms than some existing work

The rest of the paper is organized as follows In Section 1we introduce the background of our work In Section 2we discuss some classic routing protocols and some latestresearch achievement The system model which containsnetwork and energy model is presented in Section 3 InSection 4 we describe our algorithm in detail Extensivesimulations are conducted and the experimental results arediscussed and compared in Section 5 Section 6 concludes thepaper with some future work

2 Related Work

Many researchers have paid close attention to energy efficientrouting protocols or algorithms in recent years LEACH(Low-Energy Adaptive Clustering Algorithm) is one of themost famous hierarchical protocols which were proposedabout twenty years ago [15] In LEACH there are two typesof nodes namely cluster heads (CHs) and ordinary nodes(ONs) Each ON collects data from area of interest and sendsthe data package to a closest CH Each CH take charge offusing the data it receives and then transmits the fused datato the sink node The introduction of CHs can avoid longdistance communication between ON and sink node thusmuch energy is saved However the selection of CHs is ina random way as a result of random selection So the CHdistributes unevenly inside the whole network with degradedperformance Meanwhile CHs communicate with sink nodedirectly which causes much energy dissipation

PEGASIS (Power Efficient Gathering in Sensor Informa-tion Systems) is a chain-based routing protocol for WSNs[16] In PEGASIS each sensor node only needs to transmitdata to its neighbour which is closer to the sink nodeSeveral chains could be constructed according to the greedyalgorithmand the leader of each chain takes the responsibilityto transfer the data to the sink node Due to the heavy burdenof the leaders of chains each node takes turns to be the leaderto balance the energy consumption Because of multihoppropagation long distance communication between sensornode and sink node is avoided and much energy is savedMeanwhile due to multihop propagation the delay of thenetwork is very serious and it is not suitable for delay sensitiveapplications

In [17] the authors propose an energy efficient routingprotocol using mobile sink based on clustering and it issuitable forWSNswith obstruction Mobile sink moves alongthe CHs and collects data by single hop communicationIn this protocol an efficient scheduling mechanism basedon spanning graphs is proposed to find a shortest path for

mobile sink to avoid obstacle Simulation result shows thatthe lifetime of the network is prolonged and complexity ofnetwork is reduced

In [18] the authors proposed a data collection algorithmusing mobile sink based on tree clustering This algorithmcontains three phases namely tree construction tree decom-position and subrendezvous points (SRP) selection and datacollection phases Mobile sink moves towards rendezvouspoints (RP) and SRP to collect data via single hop commu-nication After data collection in each round sensor nodeswill reselect RP and SRP Simulation result proves that theproposed algorithm performs better in terms of networklifetime and the moving path of the mobile sink is short

In [19] the authors study the event-driven applicationfor WSNs with mobile sinks In this paper each sensornode has two statuses monitoring and transmission statusWhen an event is caught the status of node changes frommonitoring to transmission and the group of active sensornodes (ASN) is constructed to forward the data to the mobilesink In large-scale WSNs the area of interest is divided intoseveral subareas due to the limit of the speed of the mobilesink Each subarea contains a mobile sink and ASNs areselected to collect and transmit data In the meantime thecontinuous and optimal trajectory (COT) can be calculatedfor the mobile sink to achieve better performance

In [20] the authors propose an energy efficient routingprotocol formobile sensor networks using a path-constrainedmobile sink This protocol is suitable for WSNs which limitson the moving path of the mobile sink such as mobile sinksdeployed in bus Source node sends the data package to thetarget node which is closest to the location where the mobilesink will arrive next time and ensure the shortest path totransmit data This kind of protocol is very suitable for delay-tolerant network and it possesses higher robustness and lowerenergy consumption

In addition some heuristic algorithms are used toimprove the performance of the network [21ndash26] In [21]the authors combine the Particle Swarm Optimization (PSO)algorithms with cluster technology to enhance the net-work lifetime In [22] the authors propose an Ant ColonyOptimization (ACO) based on clustering algorithm to findan optimal moving trajectory for mobile sink In [23]techniques like Glow-worm Swarm Optimization (GSO)clustering and mobile sink are combined together to improveenergy efficiency as well as to prolong the lifetime of wirelesssensor network

3 System Model

31 Basic Assumptions In this paper we make some basicassumptions as follows

(1) All the sensor nodes are randomly deployed and keepstatic after deployment

(2) Each sensor node has a unique ID to differ from eachother

(3) All the sensor nodes have the same initial energy andbatteries of them cannot be changed










(1)



119864119877119909 (119897) = 119897 sdot 119864119890119897119890119888 (2)





119864119905119900119905119886119897

=

119896[119864119890119897119890119888 + 119864119891119904 (119898119896 )2] + (119896 minus 1) 119864119890119897119890119888 if 119898

119896 lt 1198890119896 [119864119890119897119890119888 + 119864119898119901 (119898

119896 )4] + (119896 minus 1) 119864119890119897119890119888 if 119898119896 ge 1198890

(3)


119889119900119901119905119894119898119886119897 = 1198890 (4)

119900119901119905119894119898119886119897 ℎ119900119901 119888119900119906119899119905119904 = lceil1198981198890rceil (5)











119889 (119899119904119900119906119903119888119890 119899119891119900119903119908119886119903119889119890119903)= min 10038161003816100381610038161003816119889 (119899119904119900119906119903119888119890 119899119891119900119903119908119886119903119889119890119903) minus 11988911990011990111990511989411989811988611989710038161003816100381610038161003816

(7)






119864119905ℎ119903119890119904ℎ119900119897119889 = sum119899isin119862119894 119864119903119890119904119894119889119906119886119897119873 (8)






0 5 10 15Hops

0

1

2

3

4

5

Ener

gy co

nsum

ptio

n

m=300m=400m=500

times10-6




2 ) sdot 120572 (9)













50 100 150 2000Rounds

50

100

150

200

250

300

350

400

450

500N

odes

Aliv

e







20 30 40 50 60 70 80 90 10010Rounds

0

1

2

3

4

5

Aver

age e

nerg

y co

nsum

ptio

n

times10-4









0

2

4

6

8

10

12

14

16

18

20

Aver

age h

ops

10 20 30 40 50 60 70 80 90 1000Rounds





6 Conclusions




=01=02

=03=04

0

50

100

150

200

250

300

350

400

450

500

Nod

es A

live

50 100 150 200 250 3000Rounds



0025R05R

075RR

50 100 150 200 250 3000Rounds

0

50

100

150

200

250

300

350

400

450

500

Nod

es A

live




Data Availability


Disclosure




Acknowledgments


References

































RoboticsJournal of




VLSI Design



Shock and Vibration







Journal of



Volume 2018



Volume 2018


Journal of




SensorsJournal of



RotatingMachinery





Propagation






Hindawi


Advances in

Multimedia











(1)



119864119877119909 (119897) = 119897 sdot 119864119890119897119890119888 (2)





119864119905119900119905119886119897

=

119896[119864119890119897119890119888 + 119864119891119904 (119898119896 )2] + (119896 minus 1) 119864119890119897119890119888 if 119898

119896 lt 1198890119896 [119864119890119897119890119888 + 119864119898119901 (119898

119896 )4] + (119896 minus 1) 119864119890119897119890119888 if 119898119896 ge 1198890

(3)


119889119900119901119905119894119898119886119897 = 1198890 (4)

119900119901119905119894119898119886119897 ℎ119900119901 119888119900119906119899119905119904 = lceil1198981198890rceil (5)











119889 (119899119904119900119906119903119888119890 119899119891119900119903119908119886119903119889119890119903)= min 10038161003816100381610038161003816119889 (119899119904119900119906119903119888119890 119899119891119900119903119908119886119903119889119890119903) minus 11988911990011990111990511989411989811988611989710038161003816100381610038161003816

(7)






119864119905ℎ119903119890119904ℎ119900119897119889 = sum119899isin119862119894 119864119903119890119904119894119889119906119886119897119873 (8)






0 5 10 15Hops

0

1

2

3

4

5

Ener

gy co

nsum

ptio

n

m=300m=400m=500

times10-6




2 ) sdot 120572 (9)













50 100 150 2000Rounds

50

100

150

200

250

300

350

400

450

500N

odes

Aliv

e







20 30 40 50 60 70 80 90 10010Rounds

0

1

2

3

4

5

Aver

age e

nerg

y co

nsum

ptio

n

times10-4









0

2

4

6

8

10

12

14

16

18

20

Aver

age h

ops

10 20 30 40 50 60 70 80 90 1000Rounds





6 Conclusions




=01=02

=03=04

0

50

100

150

200

250

300

350

400

450

500

Nod

es A

live

50 100 150 200 250 3000Rounds



0025R05R

075RR

50 100 150 200 250 3000Rounds

0

50

100

150

200

250

300

350

400

450

500

Nod

es A

live




Data Availability


Disclosure




Acknowledgments


References

































RoboticsJournal of




VLSI Design



Shock and Vibration







Journal of



Volume 2018



Volume 2018


Journal of




SensorsJournal of



RotatingMachinery





Propagation






Hindawi


Advances in

Multimedia











119889 (119899119904119900119906119903119888119890 119899119891119900119903119908119886119903119889119890119903)= min 10038161003816100381610038161003816119889 (119899119904119900119906119903119888119890 119899119891119900119903119908119886119903119889119890119903) minus 11988911990011990111990511989411989811988611989710038161003816100381610038161003816

(7)






119864119905ℎ119903119890119904ℎ119900119897119889 = sum119899isin119862119894 119864119903119890119904119894119889119906119886119897119873 (8)






0 5 10 15Hops

0

1

2

3

4

5

Ener

gy co

nsum

ptio

n

m=300m=400m=500

times10-6




2 ) sdot 120572 (9)













50 100 150 2000Rounds

50

100

150

200

250

300

350

400

450

500N

odes

Aliv

e







20 30 40 50 60 70 80 90 10010Rounds

0

1

2

3

4

5

Aver

age e

nerg

y co

nsum

ptio

n

times10-4









0

2

4

6

8

10

12

14

16

18

20

Aver

age h

ops

10 20 30 40 50 60 70 80 90 1000Rounds





6 Conclusions




=01=02

=03=04

0

50

100

150

200

250

300

350

400

450

500

Nod

es A

live

50 100 150 200 250 3000Rounds



0025R05R

075RR

50 100 150 200 250 3000Rounds

0

50

100

150

200

250

300

350

400

450

500

Nod

es A

live




Data Availability


Disclosure




Acknowledgments


References

































RoboticsJournal of




VLSI Design



Shock and Vibration







Journal of



Volume 2018



Volume 2018


Journal of




SensorsJournal of



RotatingMachinery





Propagation






Hindawi


Advances in

Multimedia




0 5 10 15Hops

0

1

2

3

4

5

Ener

gy co

nsum

ptio

n

m=300m=400m=500

times10-6




2 ) sdot 120572 (9)













50 100 150 2000Rounds

50

100

150

200

250

300

350

400

450

500N

odes

Aliv

e







20 30 40 50 60 70 80 90 10010Rounds

0

1

2

3

4

5

Aver

age e

nerg

y co

nsum

ptio

n

times10-4









0

2

4

6

8

10

12

14

16

18

20

Aver

age h

ops

10 20 30 40 50 60 70 80 90 1000Rounds





6 Conclusions




=01=02

=03=04

0

50

100

150

200

250

300

350

400

450

500

Nod

es A

live

50 100 150 200 250 3000Rounds



0025R05R

075RR

50 100 150 200 250 3000Rounds

0

50

100

150

200

250

300

350

400

450

500

Nod

es A

live




Data Availability


Disclosure




Acknowledgments


References

































RoboticsJournal of




VLSI Design



Shock and Vibration







Journal of



Volume 2018



Volume 2018


Journal of




SensorsJournal of



RotatingMachinery





Propagation






Hindawi


Advances in

Multimedia









50 100 150 2000Rounds

50

100

150

200

250

300

350

400

450

500N

odes

Aliv

e







20 30 40 50 60 70 80 90 10010Rounds

0

1

2

3

4

5

Aver

age e

nerg

y co

nsum

ptio

n

times10-4









0

2

4

6

8

10

12

14

16

18

20

Aver

age h

ops

10 20 30 40 50 60 70 80 90 1000Rounds





6 Conclusions




=01=02

=03=04

0

50

100

150

200

250

300

350

400

450

500

Nod

es A

live

50 100 150 200 250 3000Rounds



0025R05R

075RR

50 100 150 200 250 3000Rounds

0

50

100

150

200

250

300

350

400

450

500

Nod

es A

live




Data Availability


Disclosure




Acknowledgments


References

































RoboticsJournal of




VLSI Design



Shock and Vibration







Journal of



Volume 2018



Volume 2018


Journal of




SensorsJournal of



RotatingMachinery





Propagation






Hindawi


Advances in

Multimedia





20 30 40 50 60 70 80 90 10010Rounds

0

1

2

3

4

5

Aver

age e

nerg

y co

nsum

ptio

n

times10-4









0

2

4

6

8

10

12

14

16

18

20

Aver

age h

ops

10 20 30 40 50 60 70 80 90 1000Rounds





6 Conclusions




=01=02

=03=04

0

50

100

150

200

250

300

350

400

450

500

Nod

es A

live

50 100 150 200 250 3000Rounds



0025R05R

075RR

50 100 150 200 250 3000Rounds

0

50

100

150

200

250

300

350

400

450

500

Nod

es A

live




Data Availability


Disclosure




Acknowledgments


References

































RoboticsJournal of




VLSI Design



Shock and Vibration







Journal of



Volume 2018



Volume 2018


Journal of




SensorsJournal of



RotatingMachinery





Propagation






Hindawi


Advances in

Multimedia




=01=02

=03=04

0

50

100

150

200

250

300

350

400

450

500

Nod

es A

live

50 100 150 200 250 3000Rounds



0025R05R

075RR

50 100 150 200 250 3000Rounds

0

50

100

150

200

250

300

350

400

450

500

Nod

es A

live




Data Availability


Disclosure




Acknowledgments


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


an enhanced pegasis algorithm with mobile sink support for...

Documents