User-Density Dependent Autonomous Clustering for MANET Based on the Laplace Equation

Abstract

Forming hierarchies is critical for the scalable route management of MANETs, so autonomous clustering of users is essential if we are to autonomously establish hierarchical structures. Particularly with regard to user movement, autonomous clustering should be able to cope with various movement characteristics, including random movement of users and movement accompanied by user density changes such as commuting. Since movement with user-density changes dramatically changes MANET topology, the conventional autonomous clustering approach fails to offer sufficiently stable cluster structures. Recent studies posit that stable cluster structures are possible even if user density varies significantly, by autonomously generating the cluster heads and then replacing them as needed. This approach is effective only when the cost of cluster head replacement is small. In this paper, we propose an autonomous distributed clustering technology based on the Laplace equation that gives stable cluster structures without changing the cluster heads.