Key-peers based topology control for unstructured P2P networks

Abstract

One of the essential problems in unstructured peer-to-peer (P2P) network is enhancing the efficiency of resource retrieval. Previous researches either have poor response speed, or generate too much network overhead. In order to reduce the traffic load and improve the availability of the sharing resources in unstructured P2P networks, a key-peers based topology control mechanism is presented in this paper. Due to inherent heterogeneity of peers in P2P network, a few peers (called key-peers) directly affect the connectivity of P2P overlay topology. Therefore, it is particularly important to explore these peers. Here, we regard P2P overlay topology as an undirected graph, and analyze the similarities and differences between cut-nodes in graph theory and key-peers in P2P networks. Then we use the related principles about cut-nodes and the reachability relationship of nodes to detect the key-peers of P2P network. Furthermore, we adjust P2P overlay topology based on these key-peers. Finally, experimental results show that compared to the original P2P overlay topology, the modified topology using our approach can perform much better in success rate and respond speed, especially for rare resources.