Scalable reputation management with trustworthy user selection for P2P MMOGs

Networked virtual environments (NVEs) such as Massively Multiplayer Online Games (MMOGs) have become very popular in recent years. However, existing client-server architectures suffer from resource constrains when the number of concurrent users increases. Research on peer-to-peer networked virtual environments (P2P NVEs) thus tries to find more scalable and affordable methods via the resource sharing of mutually cooperating clients. However, P2P approaches face the problem of misbehavior by clients that do not process game rules properly. Without server monitoring, the misbehavior could challenge the proper execution of game rules and affect a game’s normal operations. In this paper, we present REPS, a distributed reputation management system for P2P-based MMOGs that allows trustworthy clients be identified and selected to perform important tasks. The first part of REPS deals with the generation, storage, and query for peer rating or game statistics (i.e., reputation factors) that could be the basis for user trustworthiness estimation. The second part of REPS deals with how to find trustworthy users by considering the weighted impacts of each factors. We propose TuS (Trustworthy user Selection) to adjust the weights of each factors so that the more important factors would weigh more. Based on the mutual rating among users and reputation queries, REPS provides a scalable and reliable reputation mechanism that facilitates decision making from choosing trustworthy superpeers, to deciding whether to interact with