A small-world caching strategy for P2P storage system

As a fundamental problem confronting modern Peer-to-Peer applications, efficient object location without deterioration in system performance has attracted great attention recently. In P2P file storage systems, caching technique is demonstrating its power in alleviating information transfer bottlenecks, balancing overall load, and reducing bandwidth consumption. Meanwhile, the Small-world network theory originated from sociology has become a recent focus of network research. This paper presents a new caching strategy based on Small-world network technology in order to address the above problems. Through adding some probabilistic features into existing caching strategies in P2P file storage systems, the interaction between peers is supposed to approach real-world network node behavior, and the Small-world caching strategy yields improvement in system performance.