Researches And Application Of The Information Search Based On The Hybrid P2P Network

The development of the information retrieval technique experienced four stages, there are manual retrieval, machinery retrieval, photoelectric retrieval and computer retrieval, they represent different periods of the development level of the science and technology respectively, especially, the development of the network and it’s application in the information management marked the period of the information age. With the development of computer and network technology both on hardware and software, the traditional C/S mode and P2P mode can not adapt to the needs of the network data communication in current situation, the hybrid P2P technology will become an important way of data communication. At the same time, the information searching from network, which is based on the hybrid P2P technique, has become a very important research direction in the information retrieval, the study and application for it can adapt to the information management in the information age, there have important practical significance.The hybrid P2P information search is a kind of important information search, which is to adapt to the current development of science and technology. It has advantages of centralized and distributed P2P information search, which combined with the index information and flooding, and it has the ability to find the deep target quickly, while avoiding the defect about the bottleneck of the centralized and the defect about needing to traverse all nodes of the distributed.In this thesis, we proposed a search algorithm that is based on the on-line time and the accessing heat, which has solved some problems in the information search based on hybrid P2P network. The main idea is to classify the nodes according to different functions, the different types of nodes will complete various tasks, all of the nodes work collaboratively, that improves the performance of the system, and achieves optimal. Because the distribution of IP address is in accordance with the network logic order, two closed IP address may be far away by the communication distance, then we use the ID logo to solve this problem. In addition, we increase super node by using the criterion of online time, which will reduce the burden of the super node in network, we set Cache by using the criterion of accessing heat, then the node access according to nearby principle.We used hierarchical network topology in the designment, the super nodes are connected by distributed DHT structure, which improved the efficiency greatly in the network search. Finally, it is used in the digital library, through the analyses of the data, it indicates the feasibility and the efficiency of the new algorithm.