A Centralized Cluster Head Selection Scheme for Reducing Discrepancy among Clusters over WSN

Abstract

Abstract. In this paper, we present a cluster head selection scheme that considers residual energy of a node and local node density to maintain cluster size and the number of clusters, which dynamically adjusts cluster size to a recommended threshold with the ever changing network dynamics of sensor network. In previous representative clustering schemes such as LEACH and D-LEACH, cluster heads are selected with a recommended probability in a distributed manner. So, there are great deviations of the number of clusters and cluster size per cluster at every round during network lifetime. To improve these discrepancies in clusters, our proposed scheme selects cluster heads on basis of amount of residential energy of nodes in centralized manner. Finally, our cluster head selection scheme can reduce discrepancies of the cluster size and the number of clusters across network lifetime comparing with existing schemes, and can potentially prolong the network lifetime.