Incentive-Based Self-Optimization Search Algorithm In The Unstructured P2P Network

With the extensive application of P2P, Application of P2P-based rapid, including search technology is one of them. Study found that, in the P2P networks, even a small amount of files shared per node, it is very large that the number of files shared in the network, in order to fully use these distributed resources, it must be able to find them quickly and accurately, therefore, P2P networks search algorithm has become an important research topic.At present, in the aspects of study for P2P search algorithms, there are based on existing unstructured search algorithm to improve research and to DHT based structured search algorithm. Paper focuses on unstructured P2P search technology and carries out analysis and research, a detailed analysis of flooding and random walk algorithms problems, and propose own improvement programs based on random walk search.In the unstructured P2P network search algorithms, flooding-based query form to Gnutella protocol-based method is easy to deploy by its independence and become absolute advantage inquiries. But on the TTL to set too high will produce large amounts of traffic, and to overcome the problem of Random Walk, though this can reduce network overhead, but can not link between the number of walker communications, leading to redundant query, and serial-based random walk has led to greater delay. Flooding and random walk common problem is that performance will dramatically decrease in the face of selfish users . All the users using uniform of the TTL control the search and can not be differentiated in user interests, on the cooperation and contributions did not inspire.Unfortunately, the performance of search greatly depends on the cooperation forward and shared nodes, however, most of P2P users are selfish by nature. Forward and respond to queries of other users will spend their own resources, so they just download the file never contribute any resources to the P2P network, this is the so-called "free-riding" problem. Therefore, most of the file download requests are directed to a small part of the selfless nodes , which willing to share ,some of these small nodes can easily overload, which causes the so-called " tragedy of the commons". In addition, the selfish users may be waiting for release of search results when repeating the same query, so many of the search request led to network congestion. With the selfishness of these selfish nodes ,a clear goal to join the network, download resources, once that is out of network; non-malicious nature,according to the protocol forwarded message,not malicious loss packet and discard message;and rational ,all choice are for acts maximize their own interests. These actions greatly damaged the flooding and random walk search performance.The predominant search schemes in unstructured P2P systems have their problems: the existence of a large number of selfish nodes, resulting in ?ooding creates excessive traffic overhead and random walk prolongs search delay. Moreover,both use uniform Time-to-Live control for all users, which makes them vulnerable to selfish user attacks,resulting in decreased query performance. In this paper, we proposed Incentive-based Self-optimization Search algorithm, the algorithm is based on random walk, through the establishment of an incentive model, the introduction of incentives, to provide differentiated search service for selfish users and ties a user's contribution to its service level. Dynamically adjust the TTL value of the random walk and the number of walker, to adjust the search performance and network overhead according to its own conditions and the current state of the network, comprehensive constraints selfish behavior of users, reducing the overhead of the network and improving the query efficiency.Compared performance of ISS with ?ooding and random walk searches with and without selfish user behaviors. The experimental results show that ISS always has the lowest search overhead without sacrificing the hit rate.