P2p Resource Search Algorithm In Its E-commerce System

In recent years, a new network application model, P2P growing in popularity because of its scalability, high fault tolerance, etc. Resource search mechanisms as the core technology of P2P applications, and its role in P2P distributed dynamic environment such as fast to meet the user requirements to find most of the system node resources. Although in recent years in the search algorithm in P2P networks in large numbers, but there is no search algorithm which will emerge.Structured P2P networks use a distributed hash (DHT) search algorithm, it looks fast, efficient, but limited by the network topology, so its application is limited; unstructured P2P networks are widely used, mainly based on flooding Discovery mechanism and the index directory search algorithm, but the search algorithm based on flooding mechanism more prone to storm the network news, while the index directory discovery mechanism based on the search algorithm has a single point of failure and network bandwidth bottlenecks problems.To solve the above problem, we do research in the following three areas:(1) In unstructured P2P network, the network topology structure is quite uncertain, the ability of each peer is in great difference. But the traditional search algorithm in P2P network capacity of each node does not discriminate, seriously affect the scalability of the system. In this paper, node ability in heterogeneous P2P networks was deeply study, proposed a set of evaluation and classification program, which can assessment each node ability more fairer so that the sharing of information resources are passed forward according to different nodes capacity after different attenuation proportions, greatly improved scalability.(2) For the shortage of the traditional Random walker algorithm in search process engender a large amount of redundant messages and can not locate the target resource quickly, the paper according to the main ideas of "Tip of the search" ability and use node evaluation and classification program, make improvement on the traditional random walker algorithm, propose a novel search algorithm——the probability search group PSG algorithm, and gives the principle of PSG algorithm analysis, implementation steps, and implementation code. Experiments show that the algorithm obviously improves the efficiency of search resources.(3) This paper presents a P2P-based e-commerce model, and realized online trading system based on P2P e-commerce, through the data stored locally the system greatly improve system security and robustness. PSG algorithm proposed will be applied to realize the search of commodity information of this system, experiments show that the algorithm maintain a low positioning overhead while improving the search efficiency.