基于Kleinberg模型的P2P网络搜索协议

P2P network search protocol based on Kleinberg model

针对基于DHT技术的结构化P2P网络存在路由效率低和负载不均衡问题,依据Kleinberg小世界模型设计了一个结构化P2P网络协议.P2P网络由一些相互连接结点类构成,结点类之间存在长程连接和短程连接,具有一定的小世界特征,减少了查询路由步数;通过设置结点类内部结点数量的最大值,可以平衡P2P网络负载;分析了P2P网络搜索开销,基于Kleinberg小世界模型的P2P网络搜索平均传递步数存在一个上界.实验结果表明,随着网络规模的扩大,平均搜索步数呈对数函数增长;长程连接数量增多可以减少平均搜索步数,减少的趋势呈反比函数.

In order to solve such problems as low routing efficiency and load imbalance existing in structurized P2P networks based on DHT technology,a structurized P2P network protocol was designed according to Kleinberg small world model.The P2P network is composed of some node clusters which are mutually connected,and there exist either short-range links or long-range links between the node clusters.The proposed protocol has certain small world characteristics,which decreases the average path length of search routing.The P2P network load can be balanced through setting the maximum value of internal node number of node clusters.The spending of P2P network search was analyzed.An upper bound existed for the average path length of P2P network search based on Kleinberg small world model.The experimental results show that with increasing the network scale,the average search path length increases in a logarithm function mode.With increasing the number of long-range links,the average search path length can be reduced,and the decreasing trend shows as an inverse ratio function mode.