Balancing Trust and Incentive in Peer-to-Peer Collaborative System

In order to maximize resource utilization as well as providing trust management in P2P computing environments, we propose a novel framework Trust-Incentive Service Management (TIM). Having child model, club model, bid model and trust model, TIM is constructed based on role based price strategy. In this mechanism, providers set the price according to demand and supply, and consumers maximize the surplus upon budget and deadline. A weighted voting scheme is also proposed to secure the P2P system by declining the join request from malicious peers. TIM is scalable and efficient in that: (1) it is completely distributed without a central server; (2) it dynamically manages the price and service by integrating valuables of pricing, trust, and incentive. A TIM prototype has been successfully implemented in a real P2P system, CROWN. We evaluate the proposed approach through comprehensive experiments and achieve improved results in service allocation efficiency, system completion time, and aggregated resource utilization.