A mobility-based framework for adaptive clustering in wireless ad hoc networks

摘要

This paper presents a novel framework for dynamically organizing mobile nodes in wireless ad hoc networks into clusters in which the probability of path availability can be bounded. The purpose of the (/spl alpha/, t) cluster is to help minimize the far-reaching effects of topological changes while balancing the need to support more optimal routing. A mobility model for ad hoc networks is developed and is used to derive expressions for the probability of path availability as a function of time. It is shown how this model provides the basis for dynamically grouping nodes into clusters using an efficient distributed clustering algorithm. Since the criteria for cluster organization depends directly upon path availability, the structure of the cluster topology is adaptive with respect to node mobility. Consequently, this framework supports an adaptive hybrid routing architecture that can be more responsive and effective when mobility rates are low and more efficient when mobility rates are high.