TBSearch: Trust based Search in Unstructured Peer-to-peer Networks

Effectiveness of P2P systems highly depends on efficiency of their search algorithms. With existence of malicious and dishonest nodes, reputation systems are essential to evaluate the trustworthiness of participating peers and to combat subversive peer behaviors. Surprisingly majority of the available reputation systems, ignore the underlying search algorithm and assume it pre-existing. In this paper we propose TBSearch, a robust and efficient trust based search algorithm for unstructured P2P networks. TBSearch maintains limited size routing indexes combining search and trust data, to guide queries to most reputable nodes. We introduce a low overhead method of estimating peers¿ reputation scores, by aggregating partial reputations obtained from reverse query path. TBSearch is adaptable to dynamics in peer joining and leaving. Through P2P network simulation experiments, we find significant performance gains in using TBSearch.