基于Chord扩展的对等定位模型研究

Research on peer-to-peer locating model based on extended chord

为减少对等网络资源定位过程中的路由跳数,提高路由效率,在研究Chord定位模型的基础上,提出了一种基于历史资源相关度的Chord扩展定位方法。利用网络中节点请求资源的访问局部性,通过计算搜索路径上节点与所查资源的相关度,将资源副本存储在相关度高的节点上,减少相同资源搜索的路由跳数;在节点中保存友邻缓存表,提高互访频繁节点的资源定位效率,缩短平均路径长度。模拟实验表明,随着查询消息的增加,扩展后的定位模型有效减少了平均路由跳数和平均查询时延,提高了资源定位的效率。

To reduce the routing hops and improve the routing efficiency of resource location in structured peer-to-peer network, based on Chord, an extended locating method is presented. The method is based on correlation degree of historical keys. Peers searching for resources have the property of locality. Using this property and calculating the correlation between the requested resource and nodes on searching path, resource replica is stored on the nodes with high correlation and the routing hops are reduced when searching for the same resource. With storing the friendly neighbor table on each node, the efficiency of resource location is improved for searching among nodes visited each other frequently, and the average path length is reduced. The simulation results show that the extended location model can reduce the average routing hops and the average searching delay with the increasing of searching messages, and the efficiency of resource location is improved.