Metaheuristic Multilevel Heterogeneous Clustering Technique for Heterogeneous Wireless Sensor Networks

Abstract

Recent advancements in miniaturization and inter-operability of wireless technologies have enabled a variant of Wireless Sensor Network (WSN), called Heterogeneous Wireless Sensor Network (HWSN). This plays a significant role in most popular applications of WSN such as habitat monitoring, volcano monitoring, military applications, etc. Similar to WSN, power resources are of utmost importance in HWSN too. Hence, several techniques have already been proposed to achieve energy-efficient network operations in heterogeneous networks. Among all those proposed schemes, the notion of clustering has proved its worth in obtaining energy-efficient clusters throughout the network. However, most of such clustering-based schemes utilize a dynamic probabilistic approach to deal with multilevel energy heterogeneity, which consumes many nodes' energy in the cluster formation. Moreover, with the varying degree of energy heterogeneity, most probabilistic schemes underperform and fail to compete. In this work, a Metaheuristic Multilevel Heterogeneous Clustering Technique (MMHCT) for Heterogeneous Wireless Sensor Networks has been proposed. MMHCT not only reduces the energy consumption in cluster-formation over the network rounds but also distributes the load evenly among the clusters to prolong the network lifetime. The supremacy of the proposed scheme is shown through various experiments against a state-of-the-art scheme, DEBCRP (Differential Evolution Based Clustering Routing Protocol) and the results confirm gains of 50%, 29.69%, 30.22%, 65.65%, & 37.17% in terms of network lifetime under 1-level, 2-level, 3-level, 4-level, & 5-level of energy heterogeneity respectively.