Research Unstructured P2P Network Resource Search Strategy

As a new distributed computing paradigm, P2P (Peer-to-Peer) networks are applied widely in resources sharing, instant communications and collaborative work. Each peer node in P2P network has the same function which can provide services as a server and receive services as a client, direct communication and file transfer can be achieved between any two nodes. Resources in P2P network are stored in each node, an effective resources search mechanism is a key for efficient resources management. P2P models are divided into centralized model, decentralized unstructured model and decentralized structured model, the algorithms of resources search are different in different model.The main work of the thesis is as follows:(1)The background and the network model of the P2P network are described, and the resource searching algorithms in unstructured P2P network model are elaborated.(2)Divisional searching strategy based on budget in unstructured P2P network was proposed for existing unstructured P2P network search algorithm in large amount of redundant messages, the problem of low efficiency of the search guide poor and search. The entire search process segment and with TTL budget mechanisms in algorithm, it strengthen the message is forwarded oriented and effective solution to the problem of a large amount of redundant messages, poor orientation and inefficient searching.(3)To improve existing algorithms because of higher search delay and network overload, the jumping algorithm in unstructured P2P network is proposed. The main idea of the jumping algorithm in unstructured P2P network is that each node in the network to save a certain radius (hops) range neighbor nodes resource index table, according to the size of the TTL to decide whether to forward the query message or query neighbor index table. Through the establishment of the index table and jumping search mechanism, greatly reduces the search delay and maintain a smaller network load.