Research On Search Algorithm Of Unstructured P2P Network

Peer-to-Peer (P2P) computing is the critical technology for the future network, such as the network is to achieve an important component of next generation Internet. How to efficiently search resources on the P2P which is the most critical issue. Unstructured P2P networks, high dynamic, self-organization, scalability and other features, makes the unstructured P2P networks more and more popular. Unstructured P2P networks have the lack of global network nodes in the knowledge of the topology, how to locate the node resources, reduce network communication overhead between the nodes of the search to become the core issue of P2P search.In response to this core issue, this paper work and innovation are as follows:1,Gossip-based communication protocols have been applied in construct unstructured P2P overlay network, but their scopes of application have the lack of extensive empirical analysis. This paper presents a generic framework--gossip-based peer sampling service to implement a reliable and efficient sampling service. Under the framework to explore and compare eight communication Protocol.Show that (rand, rand, push) and (tail, rand, push) two kinds of communication protocol structure unstructured P2P overlay network with small world properties.2,Follows from a sociological, inspiration of histological and based on complex network theory, the peer-to-peer networks have small world characteristics, in the results of research on constructed the unstructured P2P overlay network, combine topological structural optimization and information search, paper presents a search algorithm based on interests correlation and the small world introduce query algorithm and forwarding query algorithm based on interests correlation. When the node forwards query messages priority forwarded to the high interests correlation neighbor node, in order to ensure the message forwarding efficiency, to avoid the blindness of forwarding the message and reduce communication overhead in the search query for getting a higher retrieval efficiency. Experiments show the effectiveness of the algorithm.