Mobility Based Cluster Head Selection Algorithm for Mobile Ad-Hoc Network

Abstract

Abstract —Clustering in Mobile Ad-hoc Networks (MANETs) has become a crucial research issue in modern days, because clustering can improve system performance of MANETs. As MANETs have limited battery power supply, cluster formation is costly in terms of power depletion of nodes in the cluster. This is due to the huge amount of messages transmission required during cluster formation and further data transfer. In this paper, an extensive literature survey of MANET clustering algorithm was done and lastly we proposed an algorithm for cluster head selection and cluster maintenance. The proposed algorithm uses self-organizing principle for binding a node with a cluster, which can reduce the explicit message passing in cluster maintenance. Thus, there is no need of extra message passing during further cluster maintenance. Unlike most usual methods, a cluster head in our proposed algorithm acts only as an identifying tag for cluster entity, further it performs as a typical node when it is acting as head. It does not perform any cluster maintenance function or hierarchical routing and hence is not a bottleneck to the functioning of the cluster.