A Novel Cluster-based Energy Efcient Routing inWireless Sensor Networks

Aboobeker Sidhik Koyamparambil Mammu, Ashwani Sharma, Unai Hernandez-Jayo, and Nekane SainzDeustoTech, Department of Engineering, University of Deusto, Bilbao, Spain

{aboobeker.sidhik, ashwani.sharma, unai.hernandez, nekane.sainz}@deusto.es

AbstractRecent development in electronics and wireless com-munications has enabled the improvement of low-power and low-cost wireless sensors networks. Wireless Sensor Networks(WSNs)are a combination of autonomous devices transmitting locallygathered information to a so-called sink node by using multi-hop wireless routing. One of the most important challenges inWSNs is to design energy efcient routing mechanism to increasethe network lifetime due to the limited energy capacity of thenetwork nodes. Furthermore, hot spots in a WSNs emerge aslocations under heavy trafc load. Nodes in such areas quicklydrain energy resources, leading to disconnection in networkservices. Cluster based routing algorithms in WSNs have recentlygained increased interest, and energy efciency is of particularinterest. A cluster head (CH) represents all nodes in the clusterand collects data values from them. To balance the energyconsumption and the trafc load in the network, the CH shouldbe rotated among all nodes and the cluster size should be carefullydetermined at different parts of the WSNs. In this paper, weproposed an cluster based energy efcient routing algorithm(CBER), CBER elects CH based on nodes near to the optimalcluster head distance and residual energy of the nodes. In WSNsenergy is mostly consumed for transmission and reception, itis a non linear function of transmission range.In this paper,the optimal cluster head distance which links to optimal energyconsumption is derived. In addition, residual energy is consideredin the CH election in order to increase the network lifetime.Furthermore, the energy consumption of being a CH is equallyspread among the cluster members. Performance results showCBER scheme reduces the end to end energy consumption andprolong the network lifetime of multi hop network compared tothe well-known clustering algorithms LEACH and HEED.

Keywords - MANETs, Clustering, Energy efciency,Wireless Sensor Networks, and Routing.

I. INTRODUCTIONTypical sensor nodes are able to carry out sensing, data pro-

cessing, and communicating components, making it feasiblefor a wide range of promising applications, such as: environ-mental monitoring (e.g., humidity,temperature ), disaster andhealth care areas providing relief, conferencing,le exchange,commercial applications including controlling product quality,military applications and managing inventory [1]. For thesepurposes, sensors are usually deployed densely and oper-ated autonomously. Furthermore, sensor nodes are normallybattery-powered, and left alone makes it quite challenging torecharge or replace node batteries. Hence, one of the importantproblems in WSNs is how to prolong network lifetime withconstrained energy resource. If each and every node starts totransmit and receive data in the network, great data collisionsand congestions will be experienced. Therefore, the nodes in

WSNs will run out of energy very quickly. As a result, theenergy of each sensor nodes being a major limitation.Theresearch in data centric WSNs is concentrated on selection ofnetwork architecture and energy efcient data gathering, as abackbone of routing protocol and data aggregation mechanism,plays a important role in achieving it. All aspects, includingprotocols, architecture, algorithms and circuits, must be madeenergy efcient to prolong the network lifetime of the sensornode. At routing layer, the main purpose is to determine waysfor energy efcient route and reliable forwarding of data fromthe source nodes to the sink to save energy consumption.We aims on the network level energy preservation protocolsand algorithms in this paper. Due to different inherent char-acteristics of the WSNs that differentiates from other knownnetworks such as cellular or mobile ad hoc networks makesvery challenging for designing routing protocols. Furthermore,there may also be other critically-located sensors not neces-sarily close to data sinks, which carry the burden of relayinglarge amounts of data trafc, especially when multiple high-rate routes pass through these sensors. Such nodes are usuallyfrequently chosen to be data relays by routing algorithmsand may serve a large portion of the network trafc, due totheir convenient locations. Thus, avoiding the failure of suchnodes caused by early energy depletion is critical to ensure asufciently long network lifetime.

Routing algorithms can be widely classied into two cat-egories . They are direct routing and indirect routing. Indirect routing algorithms [2], [3] each nodes in WSNs directlyforward the gathered data to the sink node. Contrarily, indirectrouting algorithms [4], [5] are those clustering algorithm thatcreates different clusters of sensor nodes in WSNs. Clusteringis suggested to WSNs due to its advantages of energy saving,network scalability and network topology stability [4][6].Furthermore, clustering technique decreases the overheads oc-curred due to communication, thereby reducing interferencesand energy consumptions among network nodes. In addition,clustering improves the efciency of data relaying by decreas-ing number of nodes required to forward data in the WSNs,using data aggregation at CHs by intra cluster communicationdecreases overall packet loses. However, clustering algorithmshave some disadvantages compared to other mechanisms, suchas additional overheads during CH selection, cluster formation,and assignment process.

Clustering through formation of hierarchical WSNs helpsin efcient use of limited energy of sensor nodes and henceimproves network lifetime. Clustering has a potential im-

2013 IEEE 27th International Conference on Advanced Information Networking and Applications

1550-445X/13 $26.00 2013 IEEEDOI 10.1109/AINA.2013.106

41

pact on WSNs applications where a huge number of adhoc sensor nodes are distributed for sensing purposes. Inconventional clustering, the nodes in WSNs is divided intosmall groups called clusters and from each cluster one nodeis selected as CH which then grant other nodes to join tocreate the cluster. During information collection phase CHgathers information from all of its cluster members and afterits aggregation forwards to the next adjacent CH and last tothe sink node. While using conventional clustering, sensornodes elected as CH are believed to absorb more energy ascompared to other cluster members, due to their long rangeor multi hop transmissions to the sink node, and may leadto an irregular energy consumption of sensor nodes in theWSNs.This problem can be avoided by efcient rotation ofCH nodes and provided a major breakthrough in cluster baseddata collection in WSNs. Furthermore, regular re-election ofCHs within clusters depending on their remaining energy is apossible way to reduce the total energy consumption of everycluster. In addition, clustering scheme must be developed forvarious application scenarios to have capabilities like scalabil-ity, transmission coverage, prolonged lifetime, robustness, andsimplicity [4][6].Furthermore, the hot-spot issue is a majorconcern around sink nodes where large amounts of data aremerged. In fact, as the hop distance to a sink decreases, thetrafc on CHs quickly intensify. Hence, there is an obviousrelationship between the hop-distance to a sink node and theamount of data that has to be forwarded. This uneven energyconsumption of the CH node can rapidly disconnect the entireWSNs if the communications are extended. Many proposalshave concentrated on this problem. However, they have theirown performance limitation. We proposed CBER algorithmaimed at prolonging the overall network lifetime and energyefciency. With our scheme, the CH election depends uponthe optimal cluster head distance and remaining energy of thesensor node.

The remainder of the paper is organized as follows. InSection 2, we continue with related work. Section 3, showsthe energy consumption model. In Section 4, we derive theoptimal cluster head distance. In Section 5, we propose thenew scheme CBER. Section 6 shows the performance of theproposed scheme and we provide conclusions in Section 7.

II. RELATED WORK

Many different approaches have been carried out to de-sign feasible WSNs. Energy conservation is crucial to theprolong the network lifetime of WSNs. Many approaches forenergy efcient routing have been proposed to reduce energyconsumption. One alternative approach to conserve energy isusing clustering technique [7]. In addition, when scalability isconsidered to be a major problem when network density is ofhundreds and thousands of nodes then clustering is consideredto be a useful technique. In various WSNs applications routingefciency is considered important for energy efciency, loadbalancing, and data fusion [8]. In this paper we are concernabout CH selection schemes and discuss some of the associ-ated schemes. In clustering, only CHs need to communicatewith the sink node via multihop communication. Low Energy

Adaptive Clustering Hierarchy [3] is a well known clusteringalgorithm in which the CH in cluster is periodically rotatedamong members to achieve energy balance. However, thisscheme showed only partial success, it needs a new clusterformation process at every section. With cluster formation,in each cluster with random probability a new CH nodeis re-elected, and from the promising CH candidates, theoptimal node should be adaptively optimized for minimumcommunication distances to the maximum number of one hopneighbors. This only produce worst suboptimal solution due tocluster re-election process, which results in the nodes to spendadditional delay and energy. In addition, LEACH requires allCHs to perform single hop transmissions to the networks sink,thus it suffers from the cost of long-range transmissions. Asa result, the CHs that are further away from the sink depletestheir energy much earlier than others.

In EARACM [9] selects some overhearing nodes as CHnodes. This scheme adopted the multi-hop transmission tofurther minimize the energy absorption. Unfortunately, thisbenet comes from sacricing the resulting transmission delayand communication overhead since each CH node has tomaintain the status of the other CH nodes. In EECS [10],during cluster formation it allocates only less number ofnodes to the clusters with longer distances to the sink node.Nevertheless, it is still based on single-hop communication tothe sink from the CHs and is not scalable to large-scale or highdensity networks. In addition, clusters farther away from thesink node is limited by small cluster size by using a weightedparameter in order to reduce the energy consumption for longdistance data transmission to sink node. Simulations resultsshow that EECS performs better compared to LEACH in termsof energy efciency. However, it needs extra knowledge aboutthe WSNs, such as communication distance between CHs andSink node. Furthermore, this extra requirement adds overheadsto all nodes due to data aggregation in the WSNs.

In [11], authors proposed a protocol called HEED wherea node uses its two parameters communication cost betweencluster members and remaining energy to probabilisticallyelect itself to become a CH. HEED is a multi hop clusteringalgorithm for WSNs. Remaining energy of each sensor node isused to probabilistically choose the rst set of CHs, as usuallyperformed in other clustering techniques. In [11], communi-cation cost between cluster members shows the node degreeor nodeas proximity to the neighbor and is main parameterthat decides whether to join the cluster. However, hot spotissue in HEED appears in areas that are close to the sink, asnodes in such areas need to relay incoming trafc from otherparts of the network. Furthermore, knowledge of the wholeWSNs is necessary to determine communication cost betweencluster members. HEED is a distributed clustering mechanismin which CH nodes are picked from the WSNs. HEEDs CHselection parameter is a hybrid of energy and communicationcost. In EEUC (Energy- Efcient Unequal Clustering) [12],the cluster sizes with longer distance from sink node is largercompared to the cluster sizes near to the sink node. This wasproposed to save energy in communication between clustermembers and CHs. However, extra overheads caused due to theadditional data aggregation in all nodes degrades the efciency

42

of the network in a multihop environment. Furthermore, theenergy of CH closer to sink node tend to deplete quickly,because they forward more data trafc compared to othersensor nodes.

In Multi-hop routing protocol with unequal clustering (MR-PUC) [13] is a distributed clustering algorithm which func-tion in three phases: cluster formation, routing between CHsand sink node, and packet transmission. Each node collectsinformation from its neighbor nodes and elects a node withmaximum remaining energy as the CH. The CHs closer tosink node have smaller cluster sizes to reduce the energyfor heavy inter-cluster transmissions. However, Inter clustermulti-hop routing formation may cause an additional overhead.Hierarchical control clustering (HCC) [14] is proposed tosave energy consumption and provide scalability. In [14], newcluster generation is started when the value of the originalcluster head drops below a certain threshold. In [14], CH iselected based on the value assigned by the weighted factorto the cluster members in the cluster. Multi-tier clusteringis formed using hierarchical control clustering. In [14], thealgorithm was proposed to achieve various goals in differentlayers, such as: each sensor node will be part of a xed numberof clusters, connected clusters, and each cluster should have amaximum and minimum size constraint. However, the entirenetwork must be traversed before it can be computed and sincethe spanning tree is a global data structure it is not a strictlylocalized routing protocol.

Algorithm for cluster establishment (ACE) [15] uses aprotocol to cluster the WSNs in to constant number of it-erations, where node degree is used as the critical parameter.In [15], one way of CH selection is when many nodes inits neighborhood do not belong to any cluster, then it electitself as a CH. Furthermore, these number of iterations wereenough to achieve a stable average cluster size. In Addition,the two functional components of ACE are generation of newclusters and migration of existing ones. However, when thecommunication cost requirements and energy consumptionsare satisfying, then its hard to detect the number of iterationsrequired to achieve it. Moreover, additional overheads arecaused due to migratory mechanism in ACE. In Power-efcientand adaptive clustering hierarchy (PEACH) [16], cluster gen-eration is performed by using overhearing characteristics ofwireless communication to avoid additional overheads andsupport adaptive multi-level clustering. In WSNs, overhearinga node can recognize the source and the sink of messagestransmitted by the neighbor nodes. In [16], it can remarkablyreduce energy consumption of each node, increase the networklifetime, and is less affected by the distribution of sensornodes compared with existing clustering protocols. In brief,total energy consumption in multihop data delivery in clusteredWSNs should be analyzed comprehensively. Such an analysisshould be based on an clustering protocol and energy-efcientdata routing that prevent using network-wide broadcasts andreduces additional overhead. Furthermore, to afrm the loadbalance in a WSNs, this trade-off between the distance tothe sink from source and the cluster size should be studiedanalytically, before setting up the network hierarchy.

III. ENERGY CONSUMPTION MODELIn this paper we use a radio model proposed in [3] as radio

energy model to measure energy consumption for proposedCBER algorithm. In [3], the radio model is a combinationof three main models: transmitter, the receiver and the poweramplier. The energy consumed by the transmitter consists oftransmitter circuitry and the power amplier, and the energyconsumed in receiver for receiving data consists of the receivercircuitry [3]. When a packet transmitted from a transmitter toa receiver, where the distance between them is , the receivedsignal power at the receiver is [17] :

() =

2

(4)2(1)

where and are respectively receiver and transmittergains. Furthermore, represents any additional losses inthe packet transmission and represents carrier wavelength.The propagation loss factor , which is typically variesbetween 2 and 4. Hence, considering = = 1, and = 1, the received signal power at the receiver is

() =

2

(4)2(2)

To receive the data packets successfully, the received signalpower at the receiver must be above a minimum thresholdpower (). Hence, the transmitter signal power at thetransmitter must be above this threshold (4)

2

2 . Theenergy absorbed by the transmitter is

= (+(4)

2

2) = (+)

(3)where = (4)

2

2, which is considered as the /

absorbed in the transmitter RF amplier, is the transmit orreceive data rate (/) of each network node, is the number of bits in the packet and is the /consumed in transmitter electronics. Hence, based on [18] theenergy absorbed per second by a sensor node in three statescan be calculated as follows:

= ( + )

= () = = (1 ) = (1 )

(4)

The time for receiving and transmitting the data trafc betweenthe cluster is denoted by and respectively. Where and are the trafc data bits transmitted and receivedrespectively. The equation (5) represents the value of and.

=

=

(5)

The amount of time spent in one second for listening to theradio channel is represented as : = 1 , (0 1), thus 0 (1 1). Considering the staticdata trafc environment, where = = , the value of is represented in equation (6) as

0 12 1 (6)

43

Fig. 1. Cluster based End-End Multihop transmission in hexagonal structure

As a result, 12 , represents the maximum amountof data that can be transmitted in each cluster per second,when nodes in the cluster dont listen to the radio environmentand spend half a second for receiving the packets and anotherhalf for transmitting the packets. In simulations, we consider = 2.5 105 [19], the maximum data that can berelayed in WSNs based on equation (6) is 1.25 105per second. Where the energy consumed for listening tothe radio environment per second is represented as ., ,and are obtained from the design characteristics of thetransceivers. From [3] and [19], the specic values of is: = 3.32 107/. When = 2 109, =2.5105, = 2.4109, and 81011//2.

IV. OPTIMAL CLUSTER HEAD DISTANCE

In an end-to-end multi-hop transmission considering a equalhexagonal cell, the best route between the source and sinknode is the direct line between them, where intermediate nodesare properly deployed (the nodes exists whenever needed).As shown in Fig. 1, the data packet is transmitted from thesource to the sink node, where is the distance between them.Assume that the distance between each cluster head is , is the number of cluster heads and is derived as =

13 , = 3 ,

=

=

3 (7)

In this paper, we consider static trafc in the network. Networkis considered to have static trafc when trafc rate followingalong the network is constant. In hexagonal cluster model, represents the side of the hexagon and the optimal clusterhead distance. While is the maximum transmission rangeof node and the range should be such that two nodes locatedanywhere in adjacent cluster should be able to transmit andreceive data. The energy consumed for the end-to-end multi-hop transmission based on the energy consumption model

2 2.5 3 3.5 40

5

10

15

20

25

30

Propagation Loss Factor

Opt

imal

Clu

ster

Size

, rop

t (m)

Fig. 2. Optimal cluster head distance variation with .

discussed earlier is: = + + ,

=( + ) + () + (1

),

=(2 + ).

(8)

= ,=

3

[2 +

(13

)].

(9)

In order to nd the minimum energy consumption, we takethe rst derivative of with respect to optimal cluster headdistance, , and let () = 0

() =

3

((

13)( 1)2 2

2

), (10)

From equation (10) we can derive the value for the optimalcluster head distance as:

=113

(2

( 1)

)1/,

=113

(2

2( 1)(4)2

)1/ (11)

By using correct transceiver parameters, equation (11) showsthat the optimal cluster head location , depends on thepropagation loss factor with values ranging from 2 to 4and the network trafc. The relationship between propagationloss factor and is shown Fig. 2. In Fig. 2, the optimalcluster head distance and the side of the cell decreases whenthe propagation loss factor increases while network trafc is constant. Furthermore, there is a trade off between thenumber of cluster heads and energy consumed in each clusterhead. Based on equation (8), when the number of clusterheads is large, the size of the side of cluster becomessmall, and then energy consumption is dominated by the xedenergy consumption of each cluster head. When the number ofcluster heads is decreases, the size of the cluster side increases,and energy consumption is dominated by the cluster headbecause the energy consumed in the transmitter amplier ofeach cluster head increases quickly.

44

V. CBER: CLUSTER BASED ENERGY EFFICIENT ROUTING

In this section, we explains the procedure of the CBERthat is proposed in this paper. The CBER employ the selforganization technique for routing and clustering of WSNs.Furthermore, here a network of homogeneous sensor nodes areconsidered. In the proposed scheme, each node has to performthe basic task of sensing the eld parameters, form datapackets, and communicate with the cluster head. Clusteringin WSNs means partitioning nodes in network into differentclusters. The network model considered in this paper is ahexagonal structure shown in Fig. 1 with source nodes andsink node. Sink node is constant and xed for each simulation.Sensor nodes are homogeneous in nature, are assigned with aunique identier and have same capability. They are able tofunction in an active or sleeping state. Nodes can use transmitpower control to change the amount of power transmittedaccording to the distance of the receiver. In the CBER, eachnodes shares information about the current energy state withonly its one hop neighbors. The nodes of CBER will be in fourdifferent modes. The four modes are described as follows.

Cluster Head: While in CH mode, it gathers and aggregateinformation from its members and sends or receive messagebetween the adjacent CHs or to sink node at regular intervals.In addition, it selects the next adjacent CH node where thedata to be transferred. In hierarchical routing protocols, CHsare responsible for gathering, aggregating and forwardingthe data to the sink. Thus, they are responsible for convey-ing the complete information of its cluster members. CHis then responsible to transmit this data towards the sink.There are two types of communication for CH, Intra clustercommunication and Inter cluster communication. Intra clustercommunication is between CH and its cluster members. Intercluster communication is between CH and its adjacent CHs.

Cluster Member: A cluster member is a member that be-longs to a particular cluster, it regularly transmits the collectedinformation to its CH.

Dead Node: This is a state in which sensor node cannotoperate anymore because its energy has been depleted com-pletely or it has broken down. The node cannot either transmitor receive the data.In addition, the node is considered to be inthis state when its residual energy () is below 0.005 .

Isolated Node: This means node doesnt have any one hopneighbors either to transmit or receive the data. In this state,node is disconnected from the network.

Operation in CBER is divided into section. We performenergy aware clustering in sections and each section resultsin selection of different CH. One section means one end-to-end packet delivery. The CH is rotated among sensors ineach section and distributes the energy consumption across thenetworks. CBER is divided into setup and steady transmissionphase. The rst phase is the setup phase, where membersare classied into certain clusters and CH is selected forparticular cluster. And the second phase is the steady setphase to transmit data. Data packets is transferred intra clusterbetween CM and CH in each partition level and inter-clusterbetween CH leaders in each partition level. In the rst phase,setup phase, the hexagonal based cluster is formed and each

members in the network is assigned to different clusters.The optimal size of each cluster side for = 2 is foundfrom Fig. 2. Then cluster is formed using optimal clusterside for mitigation the hot spot problem and balance energyconsumption. This process is to make cluster with same sizeand number of cluster Members closer to the optimal clusterhead distance. Furthermore, cluster member in each clusternds the value of equation (12). Where in equation (12), N bethe number of one hop neighbor of the sensor node, is thedistance between node and node , is the average distancefrom node to all its one hop neighbors. For different clustermembers they have different values.

CH is decided based on remaining energy and optimumcluster head distance of node to mitigate hot spot problem.The CHs are selected by a rule of best candidate, which selectsa sensor node with an optimal cluster head distance and hasmaximum remaining energy as a CH for the next section. TheCBER algorithm decides the CH node as the node (within thecluster) that minimizes equation (13). In equation (13), isthe average distance to the one neighbor nodes and is theremaining energy of the neighbors in the cluster. Intuitively,without taking into the energy balance, some sensor nodesmay be selected more frequently as the cluster head nodes,and these nodes energy may be depleted out very quicklycompared to other member nodes. Since and usedifferent units, so they should be normalized and equation(13) shows the normalized form. The default energy capacityof each node () is used to normalize to dene ; is normalized with respect to the maximumdistance between two nodes in anywhere in the cluster 2 todene . The weight function determines therelative importance placed on these two parameters. This iswhy it will be an effective way to choose the proper nodeas CH, combining each of system parameter with weightingfactors chosen according application requirements. It meansthe nodes are decided to be CH depending on combinedremaining energy () of node, and the optimal cluster headdistance. This node uses the best combination of minimumenergy needed to reach neighbors and with maximum residualenergy. Therefore, CBER is higher in concept and efciencyas will be seen.

=1

= = [1, ] (12)

(, ) =

2 (1 )

(13)

Second phase is steady transmission phase, where intra andinter cluster multi-hop CH routing happens. In intra clusterrouting, cluster member sends data to the CH in the samecluster. In inter cluster routing, the data aggregated by CHwill be forwarded to adjacent CHs and later forwarded tosink node. The CH will operate at the range of which ismaximum distance between two adjacent cells in a hexagonalstructure. In order to send neighbor information with allpossible CH candidate within its cluster because CH hassent Next CH signals to optimal cluster head distance.Furthermore, the CHs create a route and select a nal CH

45

0 500 1000 1500 2000 2500 30000

50

100

150

Time Elapsed (Sec)

Pack

ets

Rece

ived

= 0 = 0.5 = 1LEACHHEED

Fig. 3. Comparison of data received by sink node

for relaying to the sink and sends messages to the sink. TheFinal CH is the cluster head node that have next hop as sinknode. Forwarding of messages through the route can reduceenergy consumption compared to the direct transmission ofmessages from all CHs to the sink. During the creation ofroute for inter-cluster routing, the CHs carry out their dutieswhile transforming into the following three modes whilerelying on different roles.Initial mode: When Inter cluster routing phase starts,all CHsare initialized as the initial state.Route broadcasting mode: This is a mode where the signalsare broadcasted to establish an inter-cluster route.Route established mode: In this mode, routes are establishedfrom its own routes and those of its neighbors.

In this paper, we compare the CBER algorithm with theLEACH [3](Low-Energy Adaptive Clustering Hierarchy) andHEED algorithm in terms of the network lifetime and throughput.

VI. SIMULATION EVALUATIONIn this section, we evaluate the performance of our pro-

posed algorithm (CBER) using MATLAB. In simulations, weassumes an error free physical layer links and ideal MAC layer.In this paper, we consider the each nodes energy consumptionas the summation of energy consumed in the transmission andreception of data packets per section. We compare CBER withLEACH and HEED. The results obtained from simulations areaverage of several tests. The simulation parameters are givenin Table II, in which the parameters of radio model are thesame as those in [3].

TABLE IMEASURMENTS FROM SIMULATION = 0 = 0.5 = 1

Lifetime (s) 1900 1800 1045 1081 1500Packets 140 143 95 86 116

We rst obtain the optimal cluster head distance () fromequation (11) ( = 25.06 m for n = 2). When is set to

0 500 1000 1500 20000.65

0.7

0.75

0.8

0.85

0.9

0.95

1

Time Elapsed (Sec)

Frac

tion

of s

urvi

ved

node

s

= 0 = 0.5 = 1LEACHHEED

Fig. 4. Comparison of network lifetime

TABLE IIPARAMETERS OF SIMULATION

Propagation factor () 2Network length 2000

Simulation time () 0.02Number of nodes ( ) 1500

Energy capacity of nodes () 2Optimal cluster size () 25.06

Min threshold power () 8 1015Carrier frequency ( ) 2.4 104

Data Rate () 2.5 105

the optimal value, it will prolong the network lifetime. InFig. 4, its shows the network lifetime based on variousvalues of from 0 to 1. The measurements obtained fromsimulation are shown in Table I. In this paper, we denethe lifetime of the network as the time duration before afraction of nodes run out of energy, and the results from Fig 4shows the efciency of each algorithm. For instance, when thedenition of lifetime of network is 30% of nodes completelydepletes its energy, the ratio of the lifetimes of network(calculated in loops between 0 until the 30 event) undervarious algorithms is: HEED:LEACH:CBER( = 1):CBER( = 0.5):CBER( = 0) = 0.78:0.56:0.55:0.94:1. It showsthe CBER algorithm (when = 0) has better performancein network lifetime when analyzed with LEACH and HEED.Furthermore, the 30% nodes referred to above corresponds to70% fraction of survived nodes in Fig. 4. Since LEACHdoes not take into consideration of the residual energy ofnodes, the network lifetime decreases drastically comparedto the other algorithms. Therefore, data packets received bysink node using LEACH should be compared. From Fig 3,before the network disconnection with the CBER ( = 0.5)algorithm, the sink node received much more data packetsthan the other algorithms. In Fig. 3, the received data withCBER( = 0.5 and 0) increases more quickly than the othertwo protocols.

VII. CONCLUSION AND FUTURE WORKIn this paper, we proposed the cluster based energy efcient

routing algorithm (CBER) to extend the network lifetime, and

46

simulation results are compared with the previous cluster basedrouting algorithms LEACH and HEED. The proposed CBERalgorithm selects the CH node as the member (within thecluster) that minimizes the value of [ 2 (1) ]. Fur-thermore, this CH node is the node that has the best residualenergy and requires the minimum energy to be reached bythe cluster members. In addition, weight parameter decidesthe relative importance placed on these two parameters. Theresults from simulations show that the CBER algorithm hasbest efciency in terms of both data packets received by sinknode and the network lifetime.

CBER creates additional overhead of control packets duringthe end-to-end packet transmission and unbalanced utilizationof nodes near sink. Our next step is to improve clustering algo-rithm to minimize the overhead of control packets and efcientutilization of nodes near sink. Furthermore, to implement indynamic trafc scenario with adjustable hexagonal structurebased on the cluster size.

REFERENCES

[1] I.F. Akyildiz, W. Su, Y. Sankarasubramaniam, and E. Cayirci, Wirelesssensor networks: a survey, Computer Networks, vol. 38, no. 4, pp. 393 422, 2002.

[2] Aboobeker, Wei Feng, and Jaafar M.H Elmirghani, Energy-efcient ge-ographic routing in ad-hoc wireless networks, London CommunicationSymposium, Sept 2009.

[3] W. R. Heinzelman and A. C. H. Balakrishnan, Energy-ef A A Acientcommunication protocol for wireless microsensor networks, in Pro-ceedings of HICSSa A A00, vol. 2, pp. 4 7, jan 2000.

[4] S. Hongchi B. Yin and Y. Shang, Analysis of energy consumption inclustered wireless sensor networks, in Wireless Pervasive Computing,2nd International Symposium on,, Feb 2007.

[5] H. Chan and A. Perrig, Ace: An emergent algorithm for highly uniformcluster formation, Proc. 1st Euro. Wksp. Sensor Networks,, pp. 154 71,Jan 2004.

[6] A. Manjeshwar and D. P. Agarwal, Apteen: A hybrid protocol forefcient routing and comprehensive information retrieval in wirelesssensor networks, Proc. Intl. Parallel and Distrib.Proc. Symp, pp. 195202.

[7] J.N. Al-Karaki and A.E. Kamal, Routing techniques in wireless sensornetworks: a survey, Wireless Communications, IEEE, vol. 11, no. 6,pp. 6 28, dec. 2004.

[8] C. Intanagonwiwat, R. Govindan, D. Estrin, J. Heidemann, and F. Silva,Directed diffusion for wireless sensor networking, Networking,IEEE/ACM Transactions on, vol. 11, no. 1, pp. 2 16, feb 2003.

[9] Sangho Yi, Geunyoung Park, Junyoung Heo, Jiman Hong, Gwangil Jeon,and Yookun Cho, Energy aware routing based on adaptive clusteringmechanism for wireless sensor networks, pp. 11151124, 2005.

[10] Guihai Chen Mao Ye, Chengfa Li and Wu.J., Eecs: an energyefcient clustering scheme in wireless sensor networks, Performance,Computing, and Communications Conference,IPCCC 2005, pp. 535 540, april 2005.

[11] O. Younis and S. Fahmy, Heed: a hybrid, energy-efcient, distributedclustering approach for ad hoc sensor networks, Mobile Computing,IEEE Transactions on, vol. 3, no. 4, pp. 366 379, oct.-dec. 2004.

[12] Chengfa Li, Mao Ye, Guihai Chen, and Jie Wu, An energy-efcientunequal clustering mechanism for wireless sensor networks, in MobileAdhoc and Sensor Systems Conference, IEEE International Conferenceon, nov. 2005, p. 8604.

[13] Bencan Gong, Layuan Li, Shaorong Wang, and Xuejun Zhou, Multihoprouting protocol with unequal clustering for wireless sensor networks,in Computing, Communication, Control, and Management, 2008. CCCM08. ISECS International Colloquium on, aug. 2008, vol. 2, pp. 552 556.

[14] S. Banerjee and S. Khuller, A clustering scheme for hierarchical controlin multi-hop wireless networks, vol. 2, pp. 1028 1037 vol.2, 2001.

[15] H. Chan and A. Perrig, Ace: an emergent algorithm for highly uniformcuster formation, Lecture Notes in Computer Science, 2004.

[16] Sangho Yi, Junyoung Heo, Yookun Cho, and Jiman Hong, Peach:Power-efcient and adaptive clustering hierarchy protocol for wirelesssensor networks, Computer Communications, vol. 30, pp. 2842 2852,2007.

[17] P.M Shankar, Introduction to Wireless Systems, John Wiley and Sons,2001.

[18] M. Bhardwaj, T. Garnett, and A.P. Chandrakasan, Upper bounds onthe lifetime of sensor networks, vol. 3, pp. 785 790 vol.3, 2001.

[19] Bolian Yin, Hongchi Shi, and Yi Shang, Analysis of energy consump-tion in clustered wireless sensor networks, feb. 2007.

47