Abstract--
The Wireless Sensor Networks consists of a largenumber of small and cheap sensor nodes that have very restrictedenergy, processing power and storage. They usually examine areas,collect data and report to the base station (BS). Due to theachievement in low-power digital circuit and wirelesscommunication, many applications of the WSN are developed andalready been used in habitat monitoring, military object and objecttracking. The disadvantage in this monitoring leads to clustering thenetworks. The hierarchal network structures which are created byclustering technique are called clusters. Clusterhead is elected by itsnearest networks. Clusterhead selection becomes a significant problem because of its dynamic environment. In this paper, the problem of suitable clusterhead selection in wireless sensor networksis analyzed. Appropriate cluster-head node election can significantlyreduce the energy consumption and enhance the lifetime of thenetwork. The fuzzy logic technique for clusterhead selection is proposed in this paper based on three descriptors, namely, Energy,Concentration and Density. The experimental results shows thesubstantial increase in the network lifetime depends on network configuration as compared to probabilistically selecting the nodes ascluster-heads using only local information.
Keywords
—
Wireless Sensor Networks, Fuzzy Logic, sensor networks, Cluster head
I.
I
NTRODUCTION
IRELESS sensor networks (WSN) are composedof a compilation of devices that communicate witheach other over a wireless medium. Such a kind of sensor network forms spontaneously whenever devicesare in transmission range. Joining and leaving of nodesoccurs dynamically, particularly when they are likemobile devices. Potential applications of wireless sensor networks can be found in traffic scenarios, ubiquitousInternet access, collaborative work, and many more.Wireless sensor networks assemble and processenvironmental data. They consist of small devicescommunicating through radio. Normally, data processing in Wireless Sensor Networks occurs locallyand decentralized. The architecture of the model isshown in Figure 1.In wireless sensor networks [5] [7], clustering is oneof the mainly popular techniques for locality-preservingnetwork organization. Cluster-based architectureseffectively decrease energy consumption, and enableefficient realization and routing protocols, dataaggregation, and security mechanisms.A cluster [6] [17] is a collection of interconnectednodes with a dedicated node called clusterhead.Clusterheads are accountable for cluster management,such as scheduling of the medium access, disseminationof control messages, or data aggregation. Therefore, therole of the clusterhead is critical for the appropriatenetwork operation. Failure of a clusterhead leads toexpensive clusterhead re-election and re-clusteringoperations.In stagnant networks, the role of the clusterhead may be assigned to any node in the cluster in a self-organizedway. Often, this role is assigned in turn to the nodes inorder to ensure fairness, as a clusterhead consumes moreenergy than a regular sensor node. An essential criterionfor the clusterhead selection is the remaining energylevel of the node. However, for fault-tolerant clusterheadselection in dynamic networks, some additional criteriafor choosing a clusterhead are required. For example,considering node mobility, if a clusterhead is close to thenetwork partition border, it may disappear from thecluster earlier than a more centrally located node. On theother hand, a centrally located node should not beselected as a clusterhead if its failure leads to cluster partitioning.The energy utilization can be minimized by allowingonly a portion of the nodes, which called cluster heads,to communicate with the base station. The data sent byeach node is then composed by cluster heads andcompressed. After that the aggregated data is transmittedto the base station. Although clustering can reduceenergy consumption [8] [9], it has certain limitations.The main setback is that energy consumption isconcentrated on the cluster heads [4]. In order toovercome this demerit, the issue in cluster routing of how to distribute the energy consumption [10] must beresolved. The representative solution is LEACH (LowEnergy Adaptive Clustering Hierarchy), which is alocalized clustering method based on a probability
An Enhanced LEACH Protocol using Fuzzy Logicfor Wireless Sensor Networks
J.Rathi/Lecturer Dr.G.Rajendran/ Prof and HeadDept.of.B.Sc(CT) Dept.of MATHSK.S.Rangasamy college of technology Kongu Engg. CollegeTiruchengode-637215 Perundurai, Erode-638 052,Tamilnadu, India. Tamilnadu, India.
W
(IJCSIS) International Journal of Computer Science and Information Security,Vol. 8, No. 7, October 2010189http://sites.google.com/site/ijcsis/ISSN 1947-5500
model. The main idea of LEACH procedure is that allnodes are chosen to be the cluster heads periodically,and each period contains two stages:
•
Construction of clusters
•
Data communicationCluster heads are selected according to the probabilityof optimal cluster heads determined by the networks.After the selection of cluster heads, the clusters areconstructed and the cluster heads communicate data with base station. Because LEACH is only depend on probability model, some cluster heads may be very closeto each other and can be located in the edge of the WSN.These disorganized cluster heads could not maximizeenergy efficiency. To overcome the defects of LEACHmethodology, a cluster head election method using fuzzylogic has been introduced. This method proved that thenetwork lifetime can be efficiently prolonged by usingfuzzy variables (concentration, energy and density). Inthis proposed method, a part of energy is spent to get thedata of the three variables especially concentration anddensity. The experimental show that the proposedapproach increases the network lifetime significantlywhen compared to LEACH approach.
Fig. 1: WSN Architecture
In this paper, a method based on LEACH using FuzzyLogic to cluster heads selection is proposed based onthree variables - battery level of node, node density anddistance from base station, and this method will beintroduced based on the assumption that the WSN canget their coordinate. Although this method has the samedrawback as of Gupta’s method, it presents a better result. For a cluster, the nodes selected by the basestation are the nodes that have the higher chance to become the cluster heads using Fuzzy Logic based ontheir battery level, node density and distance.II.
R
ELATED
W
ORKS
Handy
et al.,
[1] proposed the Low Energy AdaptiveClustering Hierarchy with Deterministic Cluster-HeadSelection. This paper focuses on reducing the power [11][13] consumption of wireless microsensor networks.Therefore, a communication protocol named LEACH(Low-Energy Adaptive Clustering Hierarchy) ismodified. The author extend LEACH’s stochasticclusterhead selection algorithm by a deterministiccomponent. Depending on the network configuration anincrease of network lifetime by about 30 % can beaccomplished. Furthermore, a new approach is presentedto define lifetime of microsensor networks using threenew metrics FND (First Node Dies), HNA (Half of the Nodes Alive), and LND (Last Node Dies).W. Heinzelman
et al.,
[2] presented an Energy-efficient Communication Protocol for WirelessMicrosensor Networks. In this paper, the author looks atcommunication protocols, which can have significantimpact on the overall energy dissipation of thesenetworks. Based on the findings that the conventional protocols of direct transmission, minimum-transmission-energy, multihop routing, and static clustering may not be optimal for sensor networks, the author proposeLEACH (Low-Energy Adaptive Clustering Hierarchy), aclustering-based protocol that utilizes randomizedrotation of local cluster base stations (cluster-heads) toevenly distribute the energy load among the sensors inthe network. LEACH uses localized coordination toenable scalability and robustness for dynamic net-works,and incorporates data fusion into the routing protocol toreduce the amount of information that must betransmitted to the base station. Simulations show thatLEACH can achieve as much as a factor of 8 reductionsin energy dissipation compared with conventionalrouting protocols. In addition, LEACH is able todistribute energy dissipation evenly throughout thesensors, doubling the useful system lifetime for thenetworks we simulated.Shen
et al,
[3] suggested the Sensor Information Networking Architecture and applications; this paper introduces a sensor information networking architecture,called SINA that facilitates querying, monitoring, andtasking of sensor networks. SINA serves the role of middleware that abstracts a network of sensor nodes as acollection of massively distributed objects. SINA'sexecution environment provides a set of configuration
(IJCSIS) International Journal of Computer Science and Information Security,Vol. 8, No. 7, October 2010190http://sites.google.com/site/ijcsis/ISSN 1947-5500
and communication primitives that enable scalable andenergy-efficient organization of and interactions amongsensor objects. On top the execution environment is a programmable substrate that provides mechanisms tocreate associations and coordinate activities amongsensor nodes. Users then access information within asensor network using declarative queries, or performtasks using programming scriptIII.
M
ETHODOLOGY
In this paper the cluster-heads are elected by the basestation in each round by calculating the chance eachnode has to become the cluster-head by consideringthree fuzzy descriptors:
•
Node concentration
•
Energy level in each node
•
Node DensityIn the proposed approach, the better cluster-heads are produced by the central control algorithm in the basestation. This is because the global knowledge about thenetwork is contained in base station. In addition, basestations are many times more potent than the sensor nodes, having sufficient memory, power and storage. Inthe proposed approach energy is spent to transmit thelocation information of all the nodes to the base station.Considering WSNs are meant to be deployed over ageographical area with the main purpose of sensing andgathering information, this paper assumes that nodeshave minimal mobility, thus sending the locationinformation during the initial setup phase is sufficient.The cluster-head collects n number of k bit messagesfrom n nodes that joins it and compresses it to cn k bitmessages with c
≤
1 as the compression coefficient. Theoperation of this fuzzy cluster-head election scheme isdivided into two rounds each consisting of a setup andsteady state phase similar to LEACH. During the setup phase the cluster-heads are determined by using fuzzy[14] knowledge processing and then the cluster isorganized. In the steady state phase the cluster-headscollect the aggregated data and performs signal processing functions to compress the data into a singlesignal. This composite signal is then sent to the basestation.The radio model used here is with Eelec = 50 nJ/bit asthe energy dissipated by the radio to run the transmitter or receiver circuitry and
ε
amp
= 100 pJ/bit/m2 as theenergy dissipation of the transmission amplifier.The energy expended during transmission andreception for a k bit message to a distance d betweentransmitter and receiver node is given by:Where,
λ
is the path loss exponent and
λ
≥
2.The model of fuzzy logic control consists of afuzzifier, fuzzy rules, fuzzy inference engine, and adefuzzifier. The most commonly used fuzzy inferencetechnique called Mamdani Method is used in the proposed approach due to its simplicity. The process is performed in four steps:•
Fuzzification
of the input variables energy,concentration and density - taking the crisp inputs fromeach of these and determining the degree to which theseinputs belong to each of the appropriate fuzzy sets.•
Rule evaluation
- taking the fuzzified inputs,and applying them to the antecedents of the fuzzy rules.It is then applied to the consequent membership function(Table 1).•
Aggregation of the rule outputs
- the process of unification of the outputs of all rules.•
Defuzzification
- the input for thedefuzzification process is the aggregate output fuzzy setchance and the output is a single crisp number.During defuzzification, it finds the point where avertical line would slice the aggregate set chance intotwo equal masses. In practice, the COG (Center of Gravity) is calculated and estimated over a sample of points on the aggregate output membership function,using the following formula:Where,
μ
A
(x) is the membership function of set A.Expert knowledge is represented based on thefollowing three descriptors:•
Node Energy
- energy level available in eachnode, designated by the fuzzy variable energy,•
Node Concentration
- number of nodes presentin the vicinity, designated by the fuzzy variableconcentration,•
Node Density
– density of node in the cluster The linguistic variables used to represent the nodeenergy and node concentration, are divided into threelevels: low, medium and high, respectively, and there arethree levels to represent the node density: sparse,medium and dense respectively. The outcome torepresent the node cluster-head election chance wasdivided into seven levels: very small, small, rather small,medium, rather large, large, and very large. The fuzzyrule base currently includes rules like the following: if the energy is high and the concentration is high and thedensity is close then the node’s cluster-head electionchance is very large.Thus, 3
3
= 27 rules are used for the fuzzy rule base. Inthis paper, the triangle membership functions are used to
(IJCSIS) International Journal of Computer Science and Information Security,Vol. 8, No. 7, October 2010191http://sites.google.com/site/ijcsis/ISSN 1947-5500
Search History:
Result 00 of 00
00 results for result for
p.
An Enhanced LEACH Protocol using Fuzzy Logic for Wireless Sensor Networks
The Wireless Sensor Networks consists of a large number of small and cheap sensor nodes that have very restricted energy, processing power and storage. They usually examine areas, collect data and report to the base station…
(More)
1.5K
Reads
4
Readcasts
0
Embed Views
More From This User
Notes
Load more
