摘要
针对卫星互联网中的SDN控制器部署问题,提出一种混合粒度的SDN多控制器部署方法,基于p-中值模型设计交换机级的初始部署机制,联合优化控制器部署位置与分配关系,实现拓扑动态下的控制时延最小化,再基于粒子群算法设计请求级的重分配机制,动态调整高负载交换机的请求分配比例,实现控制平面负载均衡。实验结果显示,该方法较传统静态部署降低时延21.18%,较单一粒度动态部署提升负载均衡度8.86%,证明其可良好适应拓扑动态性和流量时空不均性。
This paper proposes a hybrid-granularity SDN multi-controller placement strategy to address the problem of SDN controller placement in satellite Internet.Firstly,based on the p-median model,a switch-level initial placement mechanism is designed to jointly optimize the positions and assignment relationships of the controllers for minimizing the control latency under topological dynamics.Then the request-level reassignment mechanism is designed based on the particle swarm optimization algorithm,where the request distribution ratio of heavy-load switches can be dynamically adjusted to achieve load balancing in the control plane.The experimental results show that this strategy reduces the control latency by 21.18%than the traditional static placement and improves the load balance performance by 8.86%over the single-granularity dynamic placement.These shows that the proposed strategy can well adapt to the topology dynamics and the spatial-temporal heterogeneity of traffic.
作者
黄婉莹
陈始茂
周晓玲
韩旭
成泓道
HUANG Wanying;CHEN Shimao;ZHOU Xiaoling;HAN Xu;CHEN Hongdao(South China University of Technology,School of Electronics and Information Engineering,Guangzhou 510641,China;The 7th Research Institute of China Electronics Technology Group Corporation,Guangzhou 510275,China)
出处
《移动通信》
2025年第6期135-141,共7页
Mobile Communications
基金
广东省科技计划项目“卫星互联网NTN一体化融合研究与验证”(2024B0101020006)。
关键词
卫星互联网
软件定义网络
控制器部署
p-中值模型
粒子群优化
satellite Internet
software defined network
controller placement
p-median model
particle swarm optimization
