A hybrid reputation model through federation of peers having analogous function

Abstract

The widespread usage of peer to peer (P2P) system is visible in all major application areas like, file sharing , high performance computing, P2P TV, P2P IP TV etc. Conversely, the foremost confront in the growth of P2P system is the dearth of an efficient scheme to handle malicious and mysterious nodes. The QoS factor of the system will be deeply reduced by freeriders. To facilitate the augmentation, the system should be able to trim down the influence of malicious nodes with a feasible, proficient and scalable reputation model. Trust value computation among peers outline the basis of reputation model. This paper recommends a hybrid reputation model through federation of peers having analogous function. The grouping of peers having analogous function is introduced to reduce the unnecessary searching. The competence of the model can be augmented by the application of following methods. Application of election algorithm with suggested modification is a solution to central point failure. The ring structure formation of peers having analogues function reduces the search time through the network. In addition to that the behavior prediction of peers is made more accurate using the Markov chain. © 2012 Springer-Verlag GmbH.