A Hierarchical Chord Algorithm Based On Physical Topology

P2P searching technology as one of the key to P2P technologies, plays an important role in the development of the whole P2P process, right now in the mainstream of P2P searching is the target with improving network flow control, load balancing, QOS, response time. P2P must achieve good, efficient and resource sharing mechanism, which could solve the problem with resources searching and the discovery of nodes. Through the searching algorithm optimizing, P2P find a more efficient way which could search the information both in time and space of nodes finding and storage in the dynamic vary and thus improve P2P system performance, improve network bandwidth utilization.In this thesis, the problems with P2P resource searching algorithm have been studied, according to small world network inspiring and based on Chord algorithm optimization of the physical network’s topology, we propose a modified algorithm called SuperChord which based on Chord algorithm. The Chord algorithm doesn’t consider physical network’s topology between nodes when choosing routing, the messages are forwarded hops based on logic network’s topology, but has nothing to do with the physical topology, however reducing the number of hops which are forwarding messages is the key to improving network performance. So SuperChord algorithm based on Chord algorithm is proposed, which add a layer of the original Chord ring by the supernodes and ensure the path between supernodes is shortest. This algorithm extract the popular hot spot data, and simultaneously find the nearest neighbor nodes in physical network. When the P2P system receive a query request, at first it will finish the query as much as possible in the supernodes, due to supernodes store major hot spots data and shortest physical network between the supernodes, so that most queries will be completed in a small amount of physical hops. If there is nothing to find the appropriate resources in the sub-ring, it will continue to lookup resources in the outer-ring in accordance with the original Chord algorithm. Based on this searching mechanism, SuperChord can effectively reduce data traffic in the network, speed up the query speed and reduce system overhead.At last, P2Psim is used as the simulator to simulate and analyze between the original Chord protocol and SuperChord protocol. Experimental results show that the system can obtain correct information, effectively reduce network flow and short the user delay time.