Heterogeneous fuzzy based clustering protocol

Abstract

Sustaining wireless sensor nodes' energy in wireless sensor networks is a very crucial issue. Wireless Sensor Networks (WSNs) is an anthology of hundred or more sensor nodes for sensing their respective vicinity. These sensor nodes have restricted battery life and their recharging is difficult. Thus, to protract the lifespan of wireless sensor networks many optimization techniques had been introduced. Clustering is popular scheme used for enhancing the lifetime of network. Low Energy Adaptive Clustering Hierarchy (LEACH) was the first hierarchical based clustering algorithm in which cluster heads are determined using probabilistic approach in a distributed manner. After that many protocols had been introduced which utilize selecting the cluster head and rotating them to balance the energy consumption, but none of these protocol consider predicted residual energy, the energy which is left behind after performing the action as a cluster head for a complete round. Here, we introduced heterogeneity in the network with the addition of fuzzy-based clustering scheme using energy prediction for appropriate cluster head selection to optimize the energy of wireless sensor node. In this work, we introduced three different types on sensor nodes on the basis on their initial energy level. In Heterogeneous Fuzzy Based Clustering Protocol (HFCP) firstly, eligible cluster head candidates are selected using Fuzzy Inference Engine (FIS) using inputs as Residual Energy (RE) and Predicted Residual Energy (PRE) and output as the Chance to be a Cluster Head (CCH), then from the eligible candidates set cluster heads are elected using probabilistic approach including remaining energy of every node and initial energy of node. Finally, the simulation outcomes show that HFCP is more efficient in terms of First Node Dead (FND), Half Node Dead (HND) compared with LEACH and LEACH-ERE.