A heuristic crossover enhanced evolutionary algorithm for clustering wireless sensor network

Abstract

In this paper, a Heuristic-Crossover Enhanced Evolutionary Algorithm for Cluster Head Selection is proposed. The algorithm uses a novel heuristic crossover operator to combine two different solutions in order to achieve a high quality solution that distributes the energy load evenly among the sensor nodes and enhances the distribution of cluster head nodes in a network. Additionally, we propose the Stochastic Selection of Inactive Nodes, a mechanism inspired by the Boltzmann Selection process in genetic algorithms. This mechanism stochastically considers coverage effect in the selection of nodes that are required to go into sleep mode in order to conserve energy of sensor nodes. The proposed selection of inactive node mechanisms and cluster head selections protocol are performed sequentially at every round and are part of the main algorithm proposed, namely the Heuristic Algorithm for Clustering Hierarchy (HACH). The main goal of HACH is to extend network lifetime of wireless sensor networks by reducing and balancing the energy consumption among sensor nodes during communication processes. Our protocol shows improved performance compared with state-of-the-art protocols like LEACH, TCAC and SEECH in terms of improved network lifetime for wireless sensor networks deployments.