基于Kademlia的P2P网络资源定位模型改进

根据基于分布式散列表(DHT)的P2P网络资源定位方法，将虚拟节点引入结构化P2P系统，在拓扑形成时充分利用网络访问的区域性和物理网络中节点的邻近性来降低访问延迟并优化路由选择。构建一种改进的基于Kademlia的P2P网络资源定位模型，提高了网络可扩展性和可管理性及网络运行效率和资源利用率。仿真结果表明，改进后的模型继承了DHT和Kademlia的优点，在路由选择、查找成功率和平均逻辑路径长度等方面的性能均优于原Kademlia模型。

Improvement of Kademlia-based P2P Network Resource Locating Model

According to Distributed Hash Table(DHT)-based P2P network resource locating methods, this paper introduces virtual node into structured P2P system and makes the most use of regionalization of network access and contiguity of physical nodes to reduce access latency and routing length, and to optimize the routing location when the topology forms. It presents an improved Kademlia-based P2P network resource locating model. The improved model is helpful to solve the problems of expandability and manageability in P2P network. It improves the efficiencies and effectiveness of network resource. Simulation tests indicate that the improved model inherits the merits of DHT and Kademlia, and it is superior to original Kademlia model in routing location, searching success ratio and average logic path length.