Location Aided Network Model And Routing Mechanism For Mobile Peer-to-Peer Network

Mobile P2P network (Mobile peer-to-peer Network, referred to as MP2P) is composed of mutual corresponding mobile devices; it is limited by battery power, wireless bandwidth as well as dynamic network topology. As the development of the wireless communications technology and the growing number of mobile users, how to run a P2P network to compute under a mobile environment becomes a continuous goal to achieve. Timely consideration about the underlying physical topology and the establishment of a robust routing strategy is crucial in the face of problems as the dynamically changing network topology, the limited transmission range of mobile devices, the low storage of the mobile devices as well as the routing short life cycle, etc.In this paper, where in MANET flooding queries generate a large number of redundant messages at the same time poor scalability problems and Pastry in the underlying physical location of the adjacent nodes in the cover layer may be very far away query routing behavior problems occur, etc. proposed location-based assisted LDPastry network model. To the overlay network and the underlying physical network to match, which aims to reduce network latency, fewer routing hops extension, reduce redundancy, while also taking into account overhead and system performance issues. The model will be the network node in accordance with the actual physical location of the proximity divided into different domains, makes the physical location near the bottom in the logical location and close proximity to cover. The new model fully taken into account differences in the performance of the node, the node is divided into cluster head, anchor nodes and ordinary nodes, the communication between the domains by cluster head of management information within the table. Paper discusses the location of the node location method, node movement, to join, exit, failures, and routing search algorithm.Another success was the Pastry LDPastry model of leaf set and neighborhood sets into one, unified as leaf collection, making leaf collection routes in ensuring the correctness of LDPastry also enhance the locality of their work. Experimental results show that, LDPastry not only inherited the Pastry's scalability, but also reduces the network traffic, reducing routing overhead position to improve the search efficiency.