PingPangRoll:一种基于乒乓映射的P2P资源发现方法

PingPangRoll:A P2P Resource Discovered Method Based on Ping-Pang Mapping

资源发现方法严重影响着P2P系统的性能.提出了一个基于概率的称为PingPangRoll的穷举搜索算法,它结合了结构化和非结构化两者的优点.系统拓扑采用松散的随机多图结构,具有十分强的可缩放性.搜索的时候首先基于生日悖论理论计算出满足用户可靠性要求需要的数据、查询复制份数,然后通过乒乓映射将查询和数据复制精确地映射到网络节点中,以实现几乎穷举的搜索,从而克服非结构化P2P网络搜索覆盖范围有限,有资源查找不到的缺点.

A resource discovered method gravely affects the performance of P2P system. This paper proposes a probabilistic and exhaustive search algorithm called PingPangRoll, which absorbs the merits of structured and unstructured P2P. The topology of system employs a relaxed random multigraph which is scalable. Firstly evaluate the necessary number of replica of data and query by the birthday paradox theory, which is required to satisfy the user＇s reliable requirement, and then map the replica of data and query onto peers in the network by Ping-Pang mapping to perform the exhaustive search. This algorithm overcomes the drawback that the search area is limited and some existing resources can not be found in the un- structured P2P.