Applying a Socially Inspired Technique (Tags) to Improve Cooperation in P2P Networks

Abstract

In this paper, we focus on the problem of maintaining significant levels of cooperation in peer-to-peer (P2P) networks of selfish adaptive peers. We propose a simple algorithm that maintains high levels of cooperation in such a network while performing the collective task of file sharing. The algorithm is adapted from novel "tag" models of cooperation that do not rely on explicit reciprocity, reputation or trust mechanisms. A sequence of three simulation models is presented-starting with an abstract model of tag-based cooperation (TagWorld) and finishing with a P2P file-sharing model (FileWorld) that puts the technique to work. From analysis of extensive computer simulations, we demonstrate the technique to be scalable, robust, and decentralized; it requires no central servers or authorities. The algorithm is relatively simple: peers do not need to store additional trust information about other nodes or to perform significant additional processing.