An Adaptive Reputation-based Trust Framework for Peer-to-Peer Applications

Abstract

In distributed P2P environments, peers (i.e., users) often have to request the services from some unfamiliar peers (i.e., resources) that could be altruistic, selfish, or even malicious. To motivate selfish peers to cooperate and minimize the risk from malicious peers, we propose an adaptive reputation-based trust framework for peer-to-peer applications. In this framework, we define a quantifiable metric trust that is calculated from the user's own view as well as other peers' view (as references) on the reputation of a peer (resource). It can be used to quantify the trustworthiness of peers and provide a measurable and trustable way of resource scheduling/selection and access control for peer-to-peer applications. We assume economic resources compete with each other to maximize their profit and design a number of strategies, for resources to determine their price. The simulation results show that the framework supports economic resources to achieve long-term high user satisfaction, differentiates malicious nodes from normal ones and encourages the resources to provide high-quality services