A novel clustering mechanism for cluster-tree networks

Xianghua Xu,Daomin Yuan,Jian Wan

DOI: https://doi.org/10.1049/cp:20080943

2008-01-01

Abstract:In this paper, we analyze the current clustering algorithms and introduce the cluster-tree network model; then benefiting from the good organization of cluster-tree networks, we propose a novel clustering mechanism that conceptually seperate the whole networks into clusters according to serveral formulas. Then we not only make use of the advantages of cluster routing but avoid the computation and memory usage for forming and maintaining clusters. We also exploit the proposed novel clustering mechanism in well-known AODV routing. Simulation results that our clustering mechanism largely reduces routing overhead and average end-to-end delay, and performs better in terms of packets delivery ratio than AODV routing without logical clusters for cluster-tree networks, which will make the real-time and dense mesh networks more efficient and scalable.