基于DHT的P2P网络资源定位模型研究

目的分析和讨论目前流行的P2P网络模型,探讨基于DHT(Distributed Hash Table)的资源定位方法,提出一种改进的kademlia模型。方法将虚拟节点引入结构化P2P系统,对P2P网络节点进行筛选以提高网络运行效率,使之能自适应地进行网络规模调整,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特性降低访问延迟和路由长度。结果模拟测试表明模型在路由选择和访问延迟方面的表现均优于原Kademlia算法。结论改进后的模型继承了DHT和Kademlia的优点,有助于改善P2P网络的可扩展性和可管理性,优化路由,减少网络开销,从而提高网络资源的利用率,适合面向Internet规模的文件共享、协同工作等P2P应用。

Research of DHT-based Peer-to-Peer network resource locating model

Aim To analyze and discuss some popularly used P2P network models,probe into DHT-based resource locating methods and present an improved Kademlia model.Methods To introduce virtual node into structured P2P system and P2P network nodes are filtered to improve efficiency of the network,so that the network is self-adapted to adjust its scale.Make the most use of regionalization of network access and contiguity of physical nodes to reduce access latency and routing length when the topology forms.Results Simulating tests indicate that the model is superior to original Kademlia in routing path and access latency.Conclusion The improved model inherits strong points of DHT and Kademlia,is helpful to solve the problems of scalability and manageability in P2P network,optimize the routing location and mitigate network spending.Consequently it improves the effectiveness of network resource.It's suitable for large-scale P2P applications such as file sharing,network cooperation and so on.