SDN中基于多指标的链路负载均衡模型
|
摘要
由于传统网络控制层和数据层的紧耦合,其路径分配算法全局性差,实时性不足,且网络流量分配不均衡,导致网络出现拥塞。针对该问题,根据SDN集中控制和透明的优势,提出一种多指标的链路负载均衡(multi-index load balancing,MI-LB)模型。实时计算源节点到目的节点的K条可用路径,基于多指标的综合评价算法对K条可用路径综合评价,选取评价最高的路径作为转发路径。实验结果表明,该链路负载均衡模型优于传统的Dijkstra路由算法,增加了网络带宽利用率,降低了网络时延,有效提高了网络的综合性能。
Because of the tight coupling between the traditional network control layer and the data layer,the path allocation algorithm has poor globality and lacks of real-time,and the distribution of network traffic is unbalanced,which leads to network congestion.To solve the problem,according to the advantages of SDN centralized control and transparency,a load balancing model based on multiple indexes was proposed.K available paths from the source node to the destination node were calculated in real time,and based on the multi-index comprehensive evaluation algorithm,the K paths were comprehensively evaluated,and the path with highest evaluation was selected as the forwarding path.Experimental results show that the load balancing model is superior to Dijkstra algorithm,and increases the network bandwidth utilization,reduces the network delay,effectively improves the comprehensive performance of the network.
Because of the tight coupling between the traditional network control layer and the data layer,the path allocation algorithm has poor globality and lacks of real-time,and the distribution of network traffic is unbalanced,which leads to network congestion.To solve the problem,according to the advantages of SDN centralized control and transparency,a load balancing model based on multiple indexes was proposed.K available paths from the source node to the destination node were calculated in real time,and based on the multi-index comprehensive evaluation algorithm,the K paths were comprehensively evaluated,and the path with highest evaluation was selected as the forwarding path.Experimental results show that the load balancing model is superior to Dijkstra algorithm,and increases the network bandwidth utilization,reduces the network delay,effectively improves the comprehensive performance of the network.
引文
[1]ZHANG Chaokun,CUI Yong,TANG Heyi,et al.State-ofthe-art survey on software-defined networking(SDN)[J].Journal of Software,2015,26(1):62-81(in Chinese).[张朝昆,崔勇,唐翯祎,等.软件定义网络(SDN)研究进展[J].软件学报,2015,26(1):62-81.]
[2]ZUO Qingyun, CHEN Ming,ZHAO Guangsong,et al.OpenFlow-based SDN technologies[J].Journal of Software,2013,24(5):1078-1097(in Chinese).[左青云,陈鸣,赵广松,等.基于OpenFlow的SDN技术研究[J].软件学报,2013,24(5):1078-1097.]
[3]WU Yuwen.Research and design of an OpenFlow-based load balancing algorithm[D].Shanghai:East China Normal University,2014(in Chinese).[吴宇文.基于OpenFlow的网络负载均衡算法的研究与设计[D].上海:华东师范大学,2014.]
[4]LONG Hui.Research on the OpenFlow-based load-balancing routing in distributed networks[D].Shanghai:Shanghai Jiao Tong University,2013(in Chinese).[龙慧.基于OpenFlow的分布式网络中负载均衡路由的研究[D].上海:上海交通大学,2013.]
[5]WU Shan.Design and implementation of a network load balancing scheme based on SDN[D].Shanghai:Fudan University,2014(in Chinese).[吴舢.一种基于SDN的网络负载均衡方案的设计与实现[D].上海:复旦大学,2014.]
[6]RYU[EB/OL].[2017-09-20].http://ryu.readthedocs.io/en/latest/.
[7]Mc Keown N,Anderson T,Balakrishnan H,et al.OpenFlow:Enabling innovation in campus networks[J].ACM SIGCOMM Computer Communication Review,2008,38(2):69-74.
[8]Open Flowswitch specication version 1.5.1[EB/OL].[2015-04-26].https://download.csdn.net/download/yangsf/8624501.
[9]sFlow-RT[EB/OL].[2014-09-27].http://www.inmon.com/products/sFlow-RT.php.
[10]Ryu:Network delay detection application[EB/OL].[2016-05-19].https://www.sdnlab.com/16859.html(in Chinese).[Ryu:网络时延探测应用[EB/OL].[2016-05-19].https://www.sdnlab.com/16859.html.]
[11]MAO Shaowu,ZHANG Huanguo,HUANG Chongchao,et al.A new fault-tolerance mechanism in communications based on K shortest path algorithm[J].Journal of Wuhan University(Natural Science Edition),2013,59(6):534-538(in Chinese).[毛少武,张焕国,黄崇超,等.改进的K最短路径算法在通信网络中的应用[J].武汉大学学报(理学版),2013,59(6):534-538.]
[12]ZHAO Mengya,LONG Zhaohua,JIANG Guiquan,et al.Research of load balancing mechanism based on OpenFlow[J].Computer Engineering and Design,2015,36(9):2356-2360(in Chinese).[赵梦亚,龙昭华,蒋贵全,等.基于OpenFlow的负载均衡机制[J].计算机工程与设计,2015,36(9):2356-2360.]
[13]ZHENG Hongyu,DENG Yinyan,HE Ruichan.The choice of target non-dimensional method in multi target synthetic evaluations[J].Pure and Applied Mathematics,2010,26(2):319-324(in Chinese).[郑宏宇,邓银燕,贺瑞缠.综合评价中数据变换方法的选择[J].纯粹数学与应用数学,2010,26(2):319-324.]
[2]ZUO Qingyun, CHEN Ming,ZHAO Guangsong,et al.OpenFlow-based SDN technologies[J].Journal of Software,2013,24(5):1078-1097(in Chinese).[左青云,陈鸣,赵广松,等.基于OpenFlow的SDN技术研究[J].软件学报,2013,24(5):1078-1097.]
[3]WU Yuwen.Research and design of an OpenFlow-based load balancing algorithm[D].Shanghai:East China Normal University,2014(in Chinese).[吴宇文.基于OpenFlow的网络负载均衡算法的研究与设计[D].上海:华东师范大学,2014.]
[4]LONG Hui.Research on the OpenFlow-based load-balancing routing in distributed networks[D].Shanghai:Shanghai Jiao Tong University,2013(in Chinese).[龙慧.基于OpenFlow的分布式网络中负载均衡路由的研究[D].上海:上海交通大学,2013.]
[5]WU Shan.Design and implementation of a network load balancing scheme based on SDN[D].Shanghai:Fudan University,2014(in Chinese).[吴舢.一种基于SDN的网络负载均衡方案的设计与实现[D].上海:复旦大学,2014.]
[6]RYU[EB/OL].[2017-09-20].http://ryu.readthedocs.io/en/latest/.
[7]Mc Keown N,Anderson T,Balakrishnan H,et al.OpenFlow:Enabling innovation in campus networks[J].ACM SIGCOMM Computer Communication Review,2008,38(2):69-74.
[8]Open Flowswitch specication version 1.5.1[EB/OL].[2015-04-26].https://download.csdn.net/download/yangsf/8624501.
[9]sFlow-RT[EB/OL].[2014-09-27].http://www.inmon.com/products/sFlow-RT.php.
[10]Ryu:Network delay detection application[EB/OL].[2016-05-19].https://www.sdnlab.com/16859.html(in Chinese).[Ryu:网络时延探测应用[EB/OL].[2016-05-19].https://www.sdnlab.com/16859.html.]
[11]MAO Shaowu,ZHANG Huanguo,HUANG Chongchao,et al.A new fault-tolerance mechanism in communications based on K shortest path algorithm[J].Journal of Wuhan University(Natural Science Edition),2013,59(6):534-538(in Chinese).[毛少武,张焕国,黄崇超,等.改进的K最短路径算法在通信网络中的应用[J].武汉大学学报(理学版),2013,59(6):534-538.]
[12]ZHAO Mengya,LONG Zhaohua,JIANG Guiquan,et al.Research of load balancing mechanism based on OpenFlow[J].Computer Engineering and Design,2015,36(9):2356-2360(in Chinese).[赵梦亚,龙昭华,蒋贵全,等.基于OpenFlow的负载均衡机制[J].计算机工程与设计,2015,36(9):2356-2360.]
[13]ZHENG Hongyu,DENG Yinyan,HE Ruichan.The choice of target non-dimensional method in multi target synthetic evaluations[J].Pure and Applied Mathematics,2010,26(2):319-324(in Chinese).[郑宏宇,邓银燕,贺瑞缠.综合评价中数据变换方法的选择[J].纯粹数学与应用数学,2010,26(2):319-324.]