OCSSP: Optimal Cluster Size Selection-based Clustering Protocol using Fuzzy Logic for Wireless Sensor Network

Fuzzy clustering generally uses more than one input parameters to determine specific output parameters for cluster head (CH) selection. The output parameters are restricted to one or two parameters for a combination of certain specific input parameters. Different combination of input parameters computes different output values. In this paper, we extend a new multi-input multi-output clustering algorithm (OCSSP) to properly utilize the node's energy and improve the longevity of the network. Rules of the fuzzy inference system are the basis for a better outcome of the network. The rules are finalized by performing many rounds of simulations. There are three input parameters and three output parameters to completely define the clustering scenario. We perform double restriction on the size of each cluster to load balance the overall network properly. The capability to bear the load by a sensor node changes rapidly with the depletion of energy. The new proposal to restrict the size of a cluster performs better CH nomination. We compare OCSSP with Low Energy Adaptive Clustering Hierarchy (LEACH), and Distributed Unequal Clustering using Fuzzy logic (DUCF) the well-known clustering algorithms and valuable to perform clustering in WSN. The extensive simulation work proves that the OCSSP achieves significant performance gain.