A clustering method based on genetic algorithm to reduce energy consumption in wireless sensor network

Abstract

Wireless Sensor Networks (WSNs) consist of huge number of sensor nodes with limited energy that are scattered in a restricted geographic region. Since WSN's nodes are very small and have limited battery power, decreasing energy consumption due to increasing network lifetime is the biggest issue in these networks. Previous studies have demonstrated that if network nodes are organized in a number of clusters, energy dissipation reduces which leads to increase in network lifetime. Establishing control over the number and place of CHs (cluster heads) and clusters size has always been a challenge. Researchers usually ignore to consider all effective factors, such as: remaining energy, distance to BS, distance to other nodes and, number and the places of CHs during clustering. In this study, we put attention into minimizing the WSN energy consumption to increasing the network lifetime by finding the best CHs. Investigating of each WSN factors on its lifetime was studied as well. By using Genetic Algorithm (GA), almost all effective factors considered for more efficient clustering. The GA's steps were based on last researcher's findings and its fitness function' formula included most WSN’s lifetime factors. Simulation results though using NS-2 showed that the proposed algorithm decreased the WSN's energy consumption which leaded to increasing the network lifetime. It improved the First Node Dies (FND) and Half of the Nodes Dies (HND) about 40% and 50%, respectively over LEACH, 30% and 37% for LEACH-C that caused to energy balance of cluster heads and leaded to longer lifetime of network. It is evident from results that our approach satisfies our expectations. Our LEACH-CGA algorithm clearly decreases the energy consumption, postpones FND and HND and increases the WSN's lifetime.