A Particle Swarm Based Network Hosts Clustering Algorithm for Peer-to-Peer Networks

In this paper, we propose a particle swarm based algorithm to cluster peer-to-peer network hosts. Previously, the clustering of network hosts are mainly according to their connectivity [?], according to the RTTs between the hosts by probing each other, or cluster randomly. In our work, the information used to cluster the network hosts are getting from the network positioning system. The algorithm used in our study to cluster the hosts is PSO based, it is fully decentralized and suitable to the environments of peer-to-peer computing. Finally, we provide experimental evaluation of our method. The experiment shows that the nearby hosts are successfully clustered and a good average latency between hosts within same cluster