Attribute-Based Partner Updating Boosts Cooperation in Social P2P Systems

Linlin Tian,Mingchu Li,Hong Yu,Xing Jin

DOI: https://doi.org/10.1109/dasc-picom-datacom-cyberscitec.2017.196

2017-11-01

Abstract:Incentive mechanisms based on networks reciprocity encourage rational peers take an active part in social Peer-to-Peer systems. While users switch altruism partners promptly, the cooperative behavior among network groups prevails in such self-organizing systems. To further analyze the effect of partner switching rules on the evolution of cooperation, we construct a spatial Prisoner's Dilemma game model and study the coevolution of behavior strategies, social relationships of individuals and their dynamic attributes. The age of a peer, as an intrinsic property relevant to its lifetime, influences the social pattern in the realistic situation that users with the similar ages often interact actively and frequently. Accordingly we propose a new mode of partner selection based on the dynamic tag, such as the system age or strategy age. Moreover, the similarity of individual ages is introduced to the strength of relationship that determines the stability of edges in the weighted graph. Simulation results indicate that the coevolution makes an effective improvement in the survival of cooperation. The dynamic rules depress the selfish behavior in the adaptive networks more than the static rule. And the strategy aging law, in particular, has a positive influence on cooperation because it favors the forming of prosocial links and cooperative clusters.