Minimizing Both the Number of Clusters and the Variation of Cluster Sizes for Mobile Ad Hoc Networks

Abstract

The research of mobile ad hoc networks has attracted a lot of attentions recently. In particular, extensive research efforts have been devoted to the design of clustering strategies to divide all nodes into a clustering architecture such that the transmission overhead for the update of routing tables after topologic al changes can be reduced. The performance of a clustering architecture has been demonstrated to be closely related to the number of its clusters. In the paper we will address the problem of reducing the number of clusters in a clustering architecture. Since the major power of a mobile ad hoc network is battery power, a cluster head may exhaust its battery power and becomes inactive. This will lead to the cluster head’s cluster becoming broken up. Basically, the power consumption rate of a cluster head is proportional to its node degree. Thus, reducing the node degree of a cluster head will extend its lifetime and improve its cluster’s stability. The second objective in this paper is to make the number of nodes in each cluster to be equal as much as possible. To sum up, in this paper, we will propose an efficient clustering algorithm to minimize the number of clusters and the variation of cluster sizes. Computer simulations show that both the number of clusters and the variation of cluster sizes generated by our clustering algorithm are less than those generated by other existing clustering algorithms that also aim to minimize the number of clusters.