论文部分内容阅读
数据中心网络中单组播树传输往往难以应付网络拥塞和故障,从而导致可靠性和网络资源利用率较低.多组播树动态切换机制虽然能够灵活地对数据传输进行均衡和故障恢复,但在传统网络中难以部署和实现.软件定义网络将控制面和数据面分离,具有良好的可控性和可编程性.在支持Open Flow的Fat-Tree数据中心网络中,提出一种基于自适应滤波预测的多组播树动态分级切换算法.该算法首先为每个组播会话建立多个备选组播树,并根据收集的网络链路状态为每个组播树计算优先级值;然后采用自适应滤波算法对优先级值进行预测,并利用预测的优先级值设计组播树的动态分级切换策略和数据分发比例.最后,通过在Mininet平台进行了仿真实验,验证了本文所提算法的可行性和性能.
Multicast tree transport in the data center network often can not cope with network congestion and failure, resulting in lower reliability and network resource utilization.Multicast multicast tree dynamic switching mechanism can flexibly balance data transmission and fault recovery, Which is difficult to deploy and implement in traditional networks.Software defined network separates the control plane from the data plane and has good controllability and programmability.In the Fat-Tree data center network supporting Open Flow, an adaptive Filtering and predicting multi-multicast tree dynamic hierarchical switching algorithm.The algorithm firstly establishes multiple candidate multicast trees for each multicast session and calculates the priority value for each multicast tree according to the collected network link state; and then The adaptive filtering algorithm is used to predict the priority value and the dynamic classification switching strategy and the data distribution ratio of the multicast tree are designed based on the predicted priority values.Finally, simulation experiments are carried out on the Mininet platform to verify the proposed algorithm Feasibility and performance.