论文部分内容阅读
针对结构化P2P网络因忽略逻辑拓扑与物理拓扑差异而导致的路由绕路问题,首先提出一种聚类算法,将物理位置紧邻的节点自组织为若干分布式聚类;然后以聚类为单位设计了关联节点物理拓扑相邻性和逻辑拓扑相邻性的节点标识方法,并且利用该方法改进了传统Kademlia模型;最后对改进模型的性能进行了理论分析和实验测试.结果表明,改进模型通过增强每一跳路由在物理拓扑中的合理性,达到了避免路由绕路的目的,路由算法性能较传统模型提升近15%.
Aiming at the route bypass problem caused by the ignorance of logical topology and physical topology difference in structured P2P networks, a clustering algorithm is proposed to self-organize the nodes adjacent to the physical location into several distributed clusters. Then, The node identification method of associating node physical topology adjacency and logical topology adjacency is designed, and the traditional Kademlia model is improved by using this method. Finally, the theoretical analysis and experimental testing of the performance of the improved model are carried out. The results show that the improved model passes Enhance the rationality of each hop route in the physical topology, and achieve the purpose of avoiding routing bypass. The routing algorithm performance is improved by nearly 15% compared with the traditional model.