An adaptive method for node revocation in peer-to-peer networks

Different solutions have been proposed by many researchers for revocation of misbehaving nodes in peer to peer networks. In particular, the focus is on developing solutions with low computation and communication costs. Two main existing solutions are quorum-based voting and self-revocation. In an adaptive method for node revocation, each node in a network individually selects the most suitable way to respond to a detected mis-behaviour. In this work, we present a decision making procedure which chooses the best revocation strategy for each node in the network, depending on the network parameters. We are tuning an existing method for adaptive node revocation in vehicular networks, for peer to peer networks. We have also developed a mechanism to identify the trusted nodes eligible to participate in the decision making procedure as well as to collect and share certain information as required by this algorithm. The importance of establishing trustworthiness of the nodes has been identified and Eigen trust algorithm is employed for implementing the same. Finally, the efficiency of the proposed method under various network conditions is studied in comparison with the existing methods by simulation. The experimental results show that our adaptive method is faster than the original voting method. The self-revocation method is similar to adaptive method in speed. The adaptive method has showed better performance than voting and self-revocation strategies in terms of correctness of decisions. The relevance of the proposed method in a cloud computing scenario is also described.