Chord-based P2p Query Methods

The peer-to-peer network (peer to peer) is one kind of the network construction uses in information sharing, in this kind of construction, various nodes not only is the network service provider - server, also is the network service applicant -workstation, namely each computer all has the same function, not advocates from it division.Since P2P have some characteristics, large-scale, dynamic state and distribute etc, then how to search resources effectively in this kind of environment is a problem that has a challenge. Currently, the most popular network structure adopted in P2P can be divided into three structure: the P2P system with concentrated catalogue, for example, Napster,eDonkey,BitTorrent, which keeping the index information on the central sever to share the resources; the P2P system with non- structure for resources search, for example Gnutella and Freenet, which take flooding as the search method; the P2P system with structure for resources search, being like Chord, CAN, the Pastry and Tapestry which use a distribute hash table(DHT) as the foundation data structure of the system.The study of this thesis is the Chord system and its topology like ring, it provide a optimize route arithmetic by using consisted hash. Further more, the Chord system provide an expansive protocol to the dynamic state P2P system which usually have node to join or exit, it store the key value in the correspond node by using consisted hash. Each node in Chord system store some keywords which quality are equal to balance by using Consisted hash, and when the node join or exit the system its relative migration could be small. Moreover each node in the Chord merely needs to know other minority route information, can accomplish the mission that information inquires.As research existing Chord arithmetic, with the some condition, the route table maintains by node will produce the information of some redundancies, such information slow down the search speed in the large scale network. So this thesis put forward a project for improve, first according to the original algorithm to establish the route information table, then the progressive scanning, discovers these redundancies route information, and deletes these information, finally according to the characteristic of the chord, discovers the isometric from this kind of ring-like topology the new route information, and add to the route table, improves this kind of insufficiency. Through the argument, this method kept original scale of the route table, speeding the speed whiles the search resources. But it brought some blemishes, while the node build up its route table, it have to find out the redundancy information and delete, then find out some new information to replace, that increased exhaust of time.