电子学报 ›› 2021, Vol. 49 ›› Issue (8): 1615-1624.DOI: 10.12263/DZXB.20190688

• 学术论文 • 上一篇    下一篇

基于成本及功耗联合优化的SDN虚拟网络映射算法

柴蓉, 谢德胜, 陈前斌   

  1. 重庆邮电大学通信与信息工程学院,重庆 400065
  • 收稿日期:2019-06-18 修回日期:2020-11-01 出版日期:2021-08-25 发布日期:2021-08-25
  • 作者简介:柴 蓉 女,1974出生.重庆邮电大学教授、博士生导师,主要研究方向为移动通信、卫星通信、软件定义网络、无线资源管理及移动性管理技术等. E-mail:chairong@cqupt.edu.cn
    谢德胜 男,1994年出生.重庆邮电大学硕士研究生,主要研究方向为移动通信技术、软件定义网络及网络虚拟化等.
    陈前斌 男,1967年出生.重庆邮电大学副校长、教授、博士生导师,主要研究方向为个人通信、移动通信、多媒体信息处理与传输等.
  • 基金资助:
    国家自然科学基金(62071078)

Cost and Power Consumption Joint Optimization Based Virtual Network Embedding Algorithm for Software-Defined Networking

Rong CHAI, De-sheng XIE, Qian-bin CHEN   

  1. School of Communication and Information Engineering,Chongqing University of Post and Telecommunications,Chongqing 400065,China
  • Received:2019-06-18 Revised:2020-11-01 Online:2021-08-25 Published:2021-08-25

摘要:

针对多个虚拟网络请求(Virtual Network Request, VNR)动态到达的网络场景,本文提出一种基于成本及功耗联合优化的软件定义网络(Software-Defined Networking, SDN)虚拟网络映射(Virtual Network Embedding, VNE)算法.在对虚拟节点及链路映射成本及功耗进行评估的基础上,建模VNE成本及功耗的代价函数,进而在满足资源需求等约束条件下,建模基于代价函数最小化的VNE模型.该优化问题为整数线性规划问题,难以直接求解;为解决此问题,提出基于时间窗的虚拟网络批处理映射策略动态处理在线请求.继而针对特定时间窗内的VNR,将其转换为虚拟节点映射子问题和虚拟链路映射子问题,并应用启发式算法对两个子问题分别进行求解,从而确定VNR映射策略.仿真结果表明,所提算法能显著减少VNE成本及功耗,提高VNR接受率.

关键词: 软件定义网络, 虚拟网络映射, 联合优化, 启发式算法

Abstract:

For the network scenario where multiple virtual network requests (VNRs) arrive dynamically, a cost and power consumption joint optimization based virtual network embedding (VNE) algorithm was proposed for software-defined networking (SDN). Based on the evaluation of the cost and power consumption required for embedding virtual nodes and links, the cost function of VNE was formulated. Under the constraints of resource requirements, the VNE problem was formulated as cost function minimization problem. A time window-based batch embedding strategy was proposed to dynamically process online VNRs. The VNR in certain time window was transformed into virtual node embedding subproblem and virtual link embedding subproblem, and the corresponding heuristic algorithms were proposed, respectively. Simulation results showed that the proposed algorithm reduced the cost and power consumption of VNRs, and improved the acceptance ratio of VNRs.

Key words: software-defined networking, virtual network embedding, joint optimization, heuristic algorithm

中图分类号: