基于P2P和Kademlia的云计算资源定位模型设计

Design of Resource Localization Model in Cloud Computing Environment Based on P2P and Kademlia

为了解决以往基于P2P的资源定位模型所具有的拓扑失配和未考虑节点异构的问题,提出了一种基于P2P和Kademlia的云资源定位模型;首先,分析了Kademlia模型的原理和不足,在此基础了设计了一个分层的改进Kademlia资源定位模型,将整个网络的拓扑结构分为超级节点组成的主网和普通节点组成的子网,然后,对节点的路由过程、节点的定位过程、节点的加入和退出过程均进行了定义和描述,最后,采用云计算仿真工作CloudSim对文中方法进行仿真,仿真结果证明文中方法具有较高的资源定位成功率和较短的平均逻辑路径长度,其平均逻辑路径长度较其它方法平均低约12.5%。

In order to solve the topological structure matching problem and not consider the different node structure in the traditional P2P model, a cloud resource localization model based on P2P and Kademlia was proposed. Firstly, the Kedemlia model was proposed and the de- fects of it were analyzed, then the improved Kedemlia model was introduced, the network was divided into main network and sub network, the node routing, node localization and adding and leaving process was defined and described. Finally, the method in this paper was simulated in CloudSim, the result shows the method in this paper has the higher successful resource localization rate and shorter average logic route length, and the logic path length is lower about 12. 5% than the other method.