基于聚类优化的SDN多域自适应管理方法  

Adaptive Management Method for SDN Multi-domain Based on Clustering Optimization

作　　者：姚蓝 兰巨龙[1] 胡涛[1] YAO Lan;LAN Julong;HU Tao(National Digital Switching System Engineering and Technology Research Center,Zhengzhou 450002,China)

机构地区：[1]国家数字交换系统工程技术研究中心

出　　处：《计算机工程》2019年第6期119-126,共8页Computer Engineering

基　　金：国家自然科学基金创新研究群体项目(61521003);国家网络空间安全专项课题(2017YFB0803204)

摘　　要：针对软件定义网络(SDN)多域网络中存在子域规划及控制器部署不合理的问题,提出一种新的SDN多域自适应管理方法。基于改进的节点聚类思想,设计SDN子域划分优化算法,依据网络中的吸引度和归属度对节点进行聚类,完成子域规划。面向子域中不同网络代价,给出控制器自适应部署算法,并通过代价加权求和得到子域网络总开销,完成控制器部署。仿真结果表明,与控制器自适应部署算法、负载均衡算法等相比,该方法网络子域划分和控制器部署更加合理,可节约23%的备份控制器空间,改善网络的综合性能。Aiming at the problem that sub-domain planning and controller deployment are unreasonable in Software Defined Networking(SDN) multi-domain networks,a new SDN multi-domain adaptive management method is proposed.Based on the improved node clustering idea,the SDN Sub-domain Dividing Optimization Algorithm(DDOA) is designed.The nodes in the network are clustered according to the degree of attraction and attribution,and the sub-domain planning is completed.For the different network costs in the sub-domain,the Controller Adaptive Deployment Algorithm(CADA) is given,and the cost of the sub-domain network is obtained by cost-weighted summation to complete the controller deployment.Simulation results show that compared with the controller adaptive deployment algorithm and load balancing algorithm,the network sub-domain division and controller deployment are more reasonable,which can save 23% backup controller space and improve the overall performance of the network.

关 键 词：软件定义网络 子域规划 节点聚类 控制器部署 自适应管理 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]