一种基于MCDDL算法的SDN多控制器部署策略研究  

Research on SDN Multi-controller Deployment Strategy Based on MCDDL Algorithm

作　　者：刘向举[1] 赵犇 方贤进[1] 徐杨洋 LIU Xiangju;ZHAO Ben;FANG Xianjin;XU Yangyang(College of Computer Science and Engineering,Anhui University of Science and Technology,Huainan 232001,China)

机构地区：[1]安徽理工大学计算机科学与工程学院,安徽淮南232001

出　　处：《湖北民族大学学报（自然科学版）》2022年第1期50-57,共8页Journal of Hubei Minzu University:Natural Science Edition

基　　金：国家自然科学基金项目(61572034);安徽省科技重大专项(18030901025)。

摘　　要：在大规模软件定义网络架构中,控制器部署位置的不合理会直接影响网络的性能.为使控制器得到合理地部署,通过对交换机与控制器之间的平均时延、最坏时延以及控制器负载均衡度进行建模,提出了一种基于时延和负载的多控制器部署算法.首先在基于密度峰值的快速聚类算法和K-means算法中引入最短传播时延矩阵对初始聚类中心进行选取并进行节点分类;然后采用模拟退火算法对分类结果进行优化得到最佳控制器部署方案;最后,搭建仿真实验环境,对该算法的各项指标进行测试.实验结果表明,与本文相关算法相比,保证控制器负载更加均衡的同时,进一步优化网络时延,平均时延降低26.7%,最坏时延降低22.2%.In a large-scale Software-Defined Network architecture, the unreasonable deployment location of the controller will directly affect the performance of the network.In order to enable the controller to be deployed reasonably, this paper proposes a multi-controller deployment algorithm based on delay and load by modeling the average delay between the switch and the controller, the worst delay, and the degree of controller load balance.First, the shortest propagation delay matrix is introduced in the Clustering by Fast Search and Find of Density Peaks algorithm and the K-means algorithm to select the initial clustering center and perform node classification;then the Simulated Annealing algorithm is used to optimize the classification results to obtain the best controller deployment Scheme;Finally, a simulation experiment environment is set up to test the various indicators of the algorithm.The experimental results show that the network delay is further optimized while ensuring the controller load is more balanced.The average delay is reduced by 26.7%,and the worst delay is reduced by 22.2%.

关 键 词：软件定义网络 多控制器部署 平均时延 最坏时延 负载均衡 

分 类 号：TP393[自动化与计算机技术—计算机应用技术]