BaRMS: A Bayesian Reputation Management Approach for P2P Systems

Abstract

Current distributed Peer-to-Peer (P2P) applications o®er a variety of°exible and convenient services through the Internet to users from di®erent geographic locations and also support enhanced communications and interactions among them. However, security and trust are the key concerns in such applications as users in such an environment are typically unknown to each other. Trust management systems aim to decrease the risks in such applications and protect benign users from malicious users. In this paper, we introduce six attack models and propose a novel Bayesian Reputation Management System (BaRMS) for P2P environments using Bayesian probability and Markov Chain theories. BaRMS handles both positive and negative feedback. Through a case study, we show that this approach is better than the existing EigenTrust framework for P2P systems. Moreover, our simulation results of a P2P¯le sharing system also show that the proposed algorithm can greatly improve the performance over a system that does not include a trust management service under various attack models. We show that our proposed Bayesian Reputation Computation Algorithm (BaRCA) performs better than the EigenTrust algorithm under various models.

CRM: A New Dynamic Cross-Layer Reputation Computation Model in Wireless Networks

“…RCMs have been widely studied in various fields of distributed networks to support secure and trustworthy communications and collaborations among participants [4][5][6][7][8][9][10][11][12].…”

“…Multi-hop WMNs are vulnerable to both external and internal attacks [12,16] and attacks may occur at different layers. In this paper, we consider the internal multi-layer attacks.…”

“…As a key scheme for managing trust, the reputation computation model (RCM) has been introduced as an effective approach to characterize and quantify nodes' behaviors for MWNs. Although a number of RCMs and reputation-based trust functions for MWNs have been proposed in the literature, all existing RCMs were based on the direct observation of layer-specifics to evaluate the node reputation, thus ignoring many key factors of reputation in other layers [4][5][6][7][8][9][10][11][12] such as node forwarding behaviors at the network layer, channel collisions at the MAC layer, channel quality measures at the physical layer and node access behaviors at the application layer. Moreover, they did not take into account the bad mouthing attack where attackers provide dishonest recommendations to frame up good parties and/or boost trust values of malicious peers.…”