基于ACO算法的SDN网络流量调度优化研究  

Research on SDN network traffic scheduling optimization based on ACO algorithm

作　　者：李维 梁巍 周策 LI Wei;LIANG Wei;ZHOU Ce(China MobileGroup Heilongjiang Co.,Ltd.,Guangzhou 510000,China;China Mobile GBA Innvation Institute CO.,Ltd.Guangzhou 510000,China)

机构地区：[1]中国移动通信集团有限公司广东分公司,广州510000 [2]中国移动粤港澳大湾区(广东)创新研究院有限公司,广州510000

出　　处：《自动化与仪器仪表》2023年第7期42-46,共5页Automation & Instrumentation

基　　金：国家科技重大专项资金《中国移动通信有限公司下一代互联网网络升级改造》(CNGI-12-01-002)。

摘　　要：传统数据中心网络存在可扩展性差,无法对数据流实时、有效调度等问题。因此,研究利用ACO算法寻找网络流量的调度优化路径,采用SND框架进行链路状态分析,并引入信息素与启发函数对算法进行收敛和稳定性的改进。本次研究通过仿真实验验证调度策略的性能,首先得出目标函数中的时延权重和链路利用率权重最佳取值分别为0.7和0.3。同时实验表明,基于ACO算法的链路负载均衡策略下,链路的利用率相比ECMP的链路利用率提高了4.6%。同时该方法调度下流量的时延降低,时延偏差在±2 ms范围内波动。研究提出的网络数据传输调度策略可以较好地解决流量调度中存在的问题,有效提高网络吞吐量以及流量传输质量。With the rapid development of the information age,the traditional data center network management can no longer meet the rapidly expanding network data traffic transmission requirements.Therefore,ACO algorithm is used to find the optimal scheduling path of network traffic,SND framework is used for link state analysis,and pheromone and heuristic function are introduced to improve the convergence and stability of the algorithm.At the same time,the performance of the scheduling strategy is verified by simulation experiments.The experiments show that the link utilization rate of the link load balancing strategy based on ACO algorithm is 4.6%higher than that of ECMP.At the same time,the delay of traffic is reduced under the scheduling of this method,and the delay deviation fluctuates within±2 ms.The proposed network data transmission scheduling strategy can better solve the problems in traffic scheduling and effectively improve network throughput and traffic transmission quality.

关 键 词：ACO算法 流量调度 SND 链路负载 

分 类 号：TP39[自动化与计算机技术—计算机应用技术] TN911.5[自动化与计算机技术—计算机科学与技术]