Agent-based Modeling for Differentiated Admission in P2P systems Using Evolutionary Game Theory Focused on Ownership Reputation

P2P systems, which are composed of unknown agents, have many problems in data sharing and creating. The main problem is concerned with the absence of the proper incentive mechanism for cooperating and feeding the systems. If the participants of P2P systems are not supported by proper incentives for data sharing and creating, the systems will starve from resources through extinction of sharers and creators[6]. This paper analyzes the Eigenvector-based reputation system with differentiated admission, which can provide proper incentives to contributors through access priority between agents in P2P systems[1]. Based on the results from multi-agent transaction modeling using evolutionary game, we present the potential risk of ownership reputation in differentiated admission system. In our model, reputation condensation, caused from ownership reputation Eigenvalue, can be critical hindrance of vitalizing P2P systems. In conclusion, the importance of balance between incentive provisioning and transaction proliferation for reputation mechanism design is suggested.