A fuzzy approach for reputation management in Bittorrent P2P network

Managing reputation for peers in Bittorrent P2P network is very much important as the nodes are anonymous to each other and mutual trust between nodes is crucial in distributed environment. In this paper we have proposed a reputation management system that generates the reputation value for each peer automatically by taking fuzzy linguistic factors after collecting the details from peers without communicating with other peers. This helps in reducing the extra communication overhead of the system. We have taken the factors like Number of files share, Upload Bandwidth and Seeding Time of a peer that affect reputation. These three factors are fuzzified, and then the fuzzy rule base is applied to get the reputation. A reputation management algorithm has been proposed to store the reputation in a central repository which helps other peers to find the reputation about any peer.