电子学报 ›› 2018, Vol. 46 ›› Issue (9): 2229-2237.DOI: 10.3969/j.issn.0372-2112.2018.09.026

• 学术论文 • 上一篇    下一篇

运营商网络中面向时延优化的服务功能链迁移重配置策略

陈卓1,2, 冯钢2, 刘蓓2, 周杨3   

  1. 1. 重庆理工大学计算机科学与工程学院, 重庆 200433;
    2. 电子科技大学通信抗干扰技术国家级重点实验室, 四川成都 710077;
    3. 奥本大学计算机科学与软件工程学院, 美国阿拉巴马州奥本 36849
  • 收稿日期:2017-08-10 修回日期:2018-05-02 出版日期:2018-09-25 发布日期:2018-09-25
  • 作者简介:陈卓 男,1980年生,副教授,博士.研究方向为网络虚拟化、数据中心网络和网络性能分析与评估.E-mail:chenzhuo@cqut.edu.cn;冯钢 男,1964年生,教授,博士生导师,研究方向为移动通信网络、网络体系结构及网络虚拟化技术.
  • 基金资助:
    国家自然科学基金(No.61631005);中央高校基本科研业务费项目(No.ZYGX2015Z005);重庆市基础与前沿研究计划项目(No.cste2013jcyjA40024)

Delay Optimization Oriented Service Function Chain Migration and Re-deployment in Operator Network

CHEN Zhuo1,2, FENG Gang2, LIU Bei2, ZHOU Yang3   

  1. 1. College of Computer Science and Engineering, Chongqing University of Technology, Chongqing 200433, China;
    2. National Key Laboratory of Science and Technology on Communications, University of Electronic Science and Technology of China, Chengdu, Sichuan 710077, China;
    3. Department of Computer Science and Software Engineering, Auburn University, Auburn 36849, USA
  • Received:2017-08-10 Revised:2018-05-02 Online:2018-09-25 Published:2018-09-25

摘要: 在引入网络功能虚拟化(NFV)后,运营商能基于通用的计算和网络平台提供更具弹性的服务.本文研究基于NFV的运营商网络中的服务功能链(SFC)的迁移重配置问题.首先分析了业务流的动态性导致的运营商网络中资源利用率下降和负载不均的问题,然后在将计算资源、网络资源和业务流端到端限制等条件纳入考虑的基础上,建立了面向业务流服务时延优化的数学模型,并证明了SFC迁移重配置问题的NP性.进一步基于遗传算法提出了一种易于部署的启发式SFC迁移重配置策略.通过详细的对比实验表明算法在端到端时延、链路带宽资源占用率和通用服务器资源占用率等关键指标上优于其它方案12.3%、10.9%和15.8%以上.这进一步说明所提出方案通过对运营商网络资源的合理调配,在保证业务流服务质量的同时有效改善了资源利用效率.

关键词: 网络功能虚拟化, 服务功能链, 运营商网络, 遗传算法

Abstract: With the introduction of network functional virtualization (NFV),operators can provide more flexible services based on general computing and network platforms.This paper investigates the problem of migration and re-deployment of service function chain (SFC) in NFV based operator network.Firstly,this paper analyzes the problem of resource under-utilization and load unbalance in the operator network caused by the dynamic of the flow.Then,based on the consideration of the conditions such as computing resources,network resources and end-to-end restrictions,the mathematical model towards the service delay optimization is formulated,and the NP property of SFC migration and re-deployment problem is proved.Furthermore,a heuristic SFC migration and re-deployment policy is proposed based on genetic algorithm.In-depth experimental result demonstrates that the proposed solution is superior to other solutions in terms of end-to-end delay,the utilization ratio of link bandwidth resource and the utilization ratio of general server by 12.3%,10.9%,and 15.8%.The results further shows that the resource utilization is effectively improved in the case of guaranteeing the quality of service for flow,through the reasonable deployment of operator network resources.

Key words: network function visualization, service function chain, operator network, genetic algorithm

中图分类号: