A Chaotic Genetic Algorithm for Wireless Sensor Networks

Abstract

Today, clustering for Sensor Node (SN) is a method in Wireless Sensor Networks (WSNs) to diminish the energy consumption of the SN by avoiding long distance communication between the SNs. This will protract the lifetime of sensor networks. However, a cluster head has to perform various tasks such as collection of data from member nodes, aggregation of the collected data, and send that data to the BS. Network load balance is a challenging issue in WSNs for the clustering schemes. Genetic algorithms (GA) with clustering schemes are implemented for better cluster formation. The GA run through again over a large no of iterations to find the optimal solution that leads to premature convergence. The chaotic GA (CGA) will solve this problem by avoiding local convergence, i.e., by choosing a chaotic map to generate the random values instead of traditional random function and improves the performance of the traditional GA. A chaotic GA (CGA) based clustering algorithm for WSNs has been proposed in the proposed work that has better convergence rate for cluster head selection and consequently improves the performance of sensor network.