Energy Efficient Clustering Protocol for Heterogeneous Wireless Sensor Network: A Hybrid Approach Using GA and K-means

A hybrid approach combining genetic algorithm(GA) and K-means algorithm, called KGA is proposed in this paper for design of clustering protocol with energy efficiency for non-homogeneous wireless sensor network. The problem of optimal clustering can be considered as a problem for searching for an optimal number of clusters in a big search space such that WSN metrics are optimized. In the proposed protocol, distance between clusters, distance within clusters and a number of cluster heads are employed to search for optimal number of clusters and cluster heads. Maximization of energy saving and lifetime of a network are the two important metrics. The KGA is designed with a hybrid approach to population initialization scheme and objective function. The superiority of the protocol over other heuristic and meta-heuristic techniques is extensively demonstrated on several parameters: energy efficiency, network life time and throughput.