For Peer-2-Peer (P2P) networks to realize their full potential the nodes they are composed of need to coordinate and cooperate, to improve the performance of the network as a whole. This requires the suppression of selfish behavior (free-riding). Existing P2P systems often assume that nodes will behave altruistically, but this has been shown to be far from the case. We outline encouraging initial results from a P2P simulation that translates and applies the properties of "tag" models (initially developed within social simulations) to tackle these issues. We find that a simple node rewiring policy, based on the tag dynamics, quickly eliminates free-riding without centralized control. The process appears highly scalable and robust.
展开▼
