Heterogeneous-Aware Distributed Clustering for Wireless Sensor Networks

Abstract

With rapidly increasing data, clustering techniques are becoming very useful tools for analyzing data in modern research. Clustering has proven to be an effective technique for data aggregation in wireless sensor networks (WSNs). Classical clustering algorithms perform all computations in one central location such as the sink, which has a full knowledge of the network. These algorithms lead to excessive energy consumption as they rely on broadcast communication. This also leads to high data delivery latency since the sink is the only node responsible for handling all network decisions. In this paper, we propose a heterogeneous-aware distributed clustering (HADC) algorithm for WSNs. The proposed algorithm takes advantage of the presence of node heterogeneity in terms of energy in order to prolong the network lifetime. We select cluster heads (CHs) based on an effective cost function that considers residual energy and node load. In addition, HADC aims to balance load in the network and reduce energy consumption of sensor nodes by making a trade-off between node degree and distance towards potential CHs. Simulations demonstrate that HADC can achieve efficient performance in terms of several metrics as compared with HEED and LEACH. Furthermore, we conduct experiments to study the effect of nodes' heterogeneity in clustered WSNs.