摘要
当前,随着云数据服务应用的普及,因不同类型的云业务应用的网络需求不同,故不能平等地对待所有类型的云业务应用。为此,提出了一种App-RS(application-aware routing scheme)路由算法。对于calss 1业务应用,该算法考虑端对端的时延及链路负载;对于calss 2业务应用,考虑延迟偏差及链路负载;对于其他业务应用,只考虑链路负载。仿真结果表明:APP-RS路由算法比CORouting路由算法在3类业务应用中的平均吞吐量分别高出9.86%、4.53%和4.07%,相比CORouting路由算法中的calss 1业务应用平均点对点延迟少25.47%,相比CORouting路由算法中的class 2业务应用平均时延偏差少49.65%。因此,APP-RS路由算法比CORouting路由算法更能满足SDN云数据中心所有class业务应用对网络的需求。
At present,the application of cloud data services is popular. Because of different network requirements of different types of cloud applications,all types of cloud applications cannot be treated equally. It proposes a App-RS( application-aware routing scheme) routing algorithm. For calss1 business applications,the algorithm considers end-to-end delay and link load,and for calss2 business applications,considering delay bias and link load for other business applications,and only consider the link load. The simulation results show that the average throughput of App-RS routing algorithm is9. 86%,4. 53% and 4. 07% higher than CORouting routing algorithm in three kinds of business applications,and the average delay is less 25. 47% than CORouting routing algorithm,and theaverage delay deviation is 49. 65% compared with the class 2 services in CORouting routing algorithm. Therefore,the App-RS routing algorithm can meet the requirements of all the class service application of SDN more than CORouting routing algorithm.
出处
《重庆理工大学学报（自然科学）》
CAS
2017年第11期172-178,共7页
Journal of Chongqing University of Technology：Natural Science
基金
湖南省教育厅科学研究项目"基于SDN的集客专线业务传送承载解决方案研究"(15C1018)
