Hybrid Backtracking Search Optimization Algorithm and K-Means for Clustering in Wireless Sensor Networks

Rapid technology evolvement in the area of wireless sensor networks (WSNs) has led to many application-specific protocols that are particularly developed to cover different fields of usage and various network scenarios. Energy efficiency is one of the apparent challenges facing WSNs which has impacted immensely on the network performance. Hence, clustering protocols that eliminate energy inefficiencies in the network is essential. As finding an optimal set of cluster heads is an NP-hard problem, the application of heuristic algorithm is required to produce good clustering. In this paper, we propose a clustering solution for WSNs using a hybrid algorithm based on Backtracking Search Optimization Algorithm (BSA) and K-Means. A fitness function that incorporates aspects such as expected energy consumption in the network and maximum intra-cluster distance is utilized to address the problem of energy efficiency. Performance comparison against well-known clustering protocols such as LEACH and LEACH-C reveals that the hybrid of BSA and K-Means clustering algorithm is able to deliver more data to the base station and extends the network lifetime.