Improved object lookup in unstructured peer-to-peer overlay networks

Abstract

Effectiveness of unstructured P2P overlays highly depends on the efficiency and scalability of their lookup algorithms. However, how to quickly discover or lookup an object or a file that a peer is looking for in a large scale P2P network with efficiency remains a challenging research problem. In this paper we dynamically select peers with high priority of having objects while reducing the overall network traffic. Our experiments have demonstrated that our method can significantly shorten the search path and result in a higher peer search performance. We achieved a success rate of 85 percentage in locating a file when the number of peers are large in the topology with a reduction of 40 percentage of search traffic. © 2012 Springer-Verlag.