Asymmetric Clustering Approaches for Enhanced Energy Efficiency in Wireless Sensor Networks

This paper focuses on clustering and selecting an appropriate cluster head in wireless sensor networks. In symmetric clustering methods, the network is divided into several equal regions, and each region will have a cluster head regardless of the number of nodes within it. However, in our method, which employs asymmetric clustering, the centrality of nodes is calculated using the Fourier operator for the genetic algorithm. Additionally, using two other criteria—energy and dispersion—the number of cluster heads in the network is dynamically and variably selected in each round. As mentioned, in most existing methods, the cluster head was either selected in a distributed manner, leading to high energy consumption, or in a centralized manner, where one node makes decisions for the entire network, resulting in high traffic on that node. If this node encounters issues, the entire network suffers as a consequence. The proposed method, utilizing a genetic algorithm, achieved up to a 54% improvement in network energy consumption compared to the LEACH algorithm.