Research On Key Technology Of Information Retrieval Based On Unstructured P2P Network

Peer-to-peer (P2P) technology has recently received a lot of attentions, and many P2P platforms have been developed, such as Napster, MSN, BT, etc. One of the most popular P2P applications is the file-sharing system. In a P2P file-sharing system, each peer performs as both a server and a client, which means they can get resources from remote sites, meanwhile, they are also providing local resources for other peers. However, how to find and locate information in a decentralized and dynamic network is a big problem for all P2P file-sharing systems.Unfortunately, existing P2P searching mechanisms are usually dissatisfied. For example, structured P2P systems are efficient but lack of actual implements on the Internet, because of their complicated structures. Unstructured P2P systems are inefficient but more popular. From various perspectives, our work focuses on how to improve retrieval efficiency of unstructured P2P file-sharing systems. In this paper, we present a new approach to P2P information retrieval, using ant colony algorithm and information recommendation services to improve the search efficiency. Ant colony algorithm was used to make routing decisions and it made searches tend to the most favorable direction. Besides, information recommendation services could raise the file-sharing level and reduce blind searches.In order to evaluate and validate our model, we built a simulated P2P application consist of a network of peer nodes; mobile agent travel through the network, making peer nodes communicate with each other. The results have shown that our searching mechanism has good performances on the search success rate, recall rate, and load balancing.