Fuzzy based Stable Clustering Protocol forHeterogeneous Wireless Sensor Networks

-In this paper a novel Fuzzy based Stable Clustering Protocol (FSCP)for Heterogeneous Wireless Sensor Network (WSN)is proposed. The major focus of protocol is on network coverage and lifetime enhancement. It is observed that not only energy of sensor nodes but also the cluster headsdistancesaffect both network coverage and network lifetime. In this protocol, fuzzy logic with cluster head update rule is implemented to improve network coverage, throughput and network lifetime under multilevel heterogeneous network. For optimum number of cluster heads, probabilities of cluster heads are decided by fuzzy logic in which fuzzy logicselects CHs which have plenty of energy and satisfy distance criteria. Along this CH update rule also removes the unfavorable scenario of no CHs.These two factors, fuzzy logic and CH update rule better enhances the network performance when it is compared with LEACH, SEP, DEEC and EDCS protocol. Through simulation, it is justified that network coverage is improved by 17.50 % than EDCS protocol.

Here N isthe number of sensor nodes, r is simulation round, R is the maximum number of round for which network simulated, a is extra energy factor, m is fraction of advance nodes, is the residual energy of sensor node for a round particular round r.
is the average residual energy, is the probability of sensor node .
is the ideal average residual energy, ∈ (0, 1) is the weighted coefficient. 1 is the previous round average residualenergy. Proposed paper is organized as follows -Section-II provides the detail ofproposed FSCPprotocol. Section-III justified the performance of FSCP protocol with LEACH, SEP, DEEC and EDCS protocol. And finally conclusion is drawn in Section-IV.

II. FSCP (FUZZY BASED STABLE CLUSTERING PROTOCOL) MODEL
A WSN with cross section M×M is deployed with N number of sensor nodes Fig.1. Under multilevel heterogeneous network, let be the initial energyof deployed node .
is the maximum energy sensor node from deployed nodes. Base station is located at center (50,50) and let the distance of normal nodes form BS are _ .If maximum distance from BS is then fuzzy inputs are given in eq. (7,8). Based on these two inputs fuzzy rule will decide which node will be more probable than others. The outcome of fuzzy input is fed as crisp input to the network and same thresholding is applied as EDCS protocol eq. (9). Fuzzy rule with its input output membership function has been indicated in Fig. 2(a, b,c andd). For CHs selection, Mamdani model [10] is used in fuzzy rule because it is acceptable worldwide for variable input parameters.Modeling of fuzzy rule is based on two factor, (i) Normalized energy of sensor nodes ( ) and (ii) normalized distance ( ). In fuzzy rule, crisp input ( (i), ) are converted into fuzzy input. These fuzzified input applied to inference rule as given in Fig. 2 (d)and the outcome of fuzzy rule will be treated as probability of CHs selection.Based ontheseprobabilities and CHs threshold eq. [9],certain nodes will undergo in CH selection process. It provides optimum solution in terms of CHs probability for each node. Since network is multilevel heterogeneous and each node has different probability, at the end stage of simulation,there may be the chance that most of the nodes which are having less probability will not undergo in CHs selection process. It creates unfavorable scenario for network. To overcome this, CH update rule is necessary for proper clustering unless optimum number of nodes (given in eq. (9) [7]) are selected as CH.Flow chart of the network is shown in Fig. 3.

III. SIMULATION PARAMETERS AND RESULTS
To compare the performance of proposed FSCP with LEACH, SEP, DEEC and EDCS protocols, same network parameters has been taken as EDCS protocol [9]. Parameters are indicated in Table-I [9]. Fig.4 (a) shows the alive node vs round. It indicates that life time of the network with stability period (17.50 % improvement over EDCS) is much better than existing protocol ([6] - [9]). Fig, 4 (b) shows the FSCP stability period comparison with ([6] - [9]) protocols. It indicates that under multilevel heterogeneous network with different energy heterogeneity parameter (lamda), FSCP stability period is much improved over EDCS. At some points in energy heterogeneity parameter (lamda2,3 and 3.5) performance of FSCP in not better but overall performance is remarkably improved. Fig. 4 (c, d) compares the die-up of alive nodes at 10%, and 50% it shows that life time FSCP is improved with network coverage. Also for different energy heterogeneity parameter (lamda), packet received at BS are much better in comparison to existing protocol ([6] - [9]) Fig. 4 (e and f).Therefore, through MATLAB simulation it is depict that overall performance of proposed FSCP is much better than other protocols ([6] - [9]). IV. CONCLUSION In this paper, FSCP (Fuzzy based Stable Clustering Protocol)for multilevel heterogeneous WSN is proposed.In FSCP,Fuzzy logic with CH update rule enhances the network coverage and network lifetime under different energy heterogeneity parameters. ThisImprovement in network coverage also enhances the throughput at BS. This is because of synchronous energy consumption of sensor nodes after employing the balance effect of normalized energy and normalized distance. Simulation results justified that overall network performance is better than existing protocols LEACH, SEP, DEEC and EDCS.