A Mobile P2P Resource Search Strategy Based On Grouping

In recent years, with the rapid popularization of the mobile wireless network and the emerging in endlessly of many kinds of mobile terminals, mobile wireless network are developing rapidly. At the same time. P2P technology are widely used in the traditional cable network, such as distributed computing, file sharing, video playback, IP layer voice calls, game network platform and so on, it almost goes to every aspects of people’s life. Therefore, more and more research institutions are getting active to the field of mobile P2P network. Mobile P2P network also called mobile peer-to-peer network, all P2P application used in wireless network can be put into this field.At present, the research to the mobile P2P mostly focus on the system structure, data distribution, resource search, security and privacy issues, reliability’transmission, etc. Among them, the hottest research is resource search algorithm.The current research of mobile P2P network resource search algorithm are mainly from two aspects:first, improving the existing P2P resource search algorithm, such as the structural search algorithm Chord, Pastry,CAN, the flooding algorithm Gnutella, Gia, to make them suitable for the mobile environment. Second, proposing a new resource search scheme to apply to the mobile P2P network. But the present study always has some problems, and it doesn’t appear a very mature scheme. The main reason is the high mobility of the mobile wireless network, the frequent alteration of the network topology, the limited storage capacity of the mobile terminal equipment, etc. They are the mobile wireless network characteristics. These characteristics make the performance of the mobile P2P network bad. the high failure rate of the nodes, the long query delay, the low data transmission efficiency, and the low efficiency in resource search.In this paper, we proposed a mobile P2P resource search strategy based on grouping, this algorithm is based on the Chord algorithm and the flooding algorithm. According to the different interests of the nodes, it divides the nodes into different group. Each group is comprised by common nodes and super nodes. Super nodes are mainly comprised by the nodes with high performance, strong storage and processing ability, stable station in the group. The connection between the group are completed by the super nodes, which form the Chord ring in the whole network. The resource sharing between the nodes in the same group are completed by the centralized directory algorithm. Therefore, the whole system forms a two-layer topology structure. Also according to the mobility of the network, the algorithm adjusts the system TTL dynamically. The simulation results shows that this algorithm in the paper shortens the file request time, improves the success rate of the resource search.