An Information Theoretic Approach to Cluster Head Selection in Mobile Ad Hoc Networks

摘要

Hierarchical routing is found to be more effective than flat routing in large mobile ad hoc networks. For this purpose, mobile nodes are grouped into clusters and each cluster is represented by a cluster head. Conventional methods use either the connectivity or the identifiers of the nodes to select the cluster head. We propose an activity based approach to select the cluster head. We use the entropy of the node to gauge the activity. The entropy is a measure of how confined the movements are, and this is used to classify nodes as conservative and exploratory. Simulation results show that the proposed cluster head selection causes less cluster head changes, as compared to the widely used Lowest-Id cluster head selection.