基于网络划分的SDN分布式控制器部署
DOI:
作者:
作者单位:

1.西南石油大学 信息学院;2.四川大学 计算机学院

作者简介:

通讯作者:

中图分类号:

基金项目:

国家“863”高技术发展计划项目(2008AA01Z105);四川省南充市科技项目(NC17SY1001, 19SXHZ0012)


Distributed controller placement in SDN based on network partitioning
Author:
Affiliation:

1.College of Information,Southwest Petroleum University;2.College of Computer Science,Sichuan University

Fund Project:

The National 863 High Technology Research and Development Program(2008AA01Z105);Science And TechnologyProject of Nanchong City(NC17SY1001, 19SXHZ0012)

  • 摘要
  • |
  • 图/表
  • |
  • 访问统计
  • |
  • 参考文献
  • |
  • 相似文献
  • |
  • 引证文献
  • |
  • 资源附件
  • |
  • 文章评论
    摘要:

    针对大规模SDN网络中分布式控制器部署问题,以优化网络弹性和可靠性为目标,提出两阶段控制器部署算法(TSCP):利用节点相似度划分控制域,使得控制域内设备之间的连通性强、连接紧密,增强控制域的网络弹性;选择控制路径平均失效率最小的控制器集合作为控制器的部署,提高网络的可靠性。通过约束控制域的规模和设备(交换机或控制器)之间传播时延,使得控制域的交换机个数均衡,控制器的部署合理。通过定义性能指标,实验对比GCP算法、K*-means算法,结果表明TSCP算法可以优化控制域的规模,均衡控制域的交换机个数,减少控制器个数,网络弹性和可靠性均表现较好。

    Abstract:

    To deploy distributed controller in large-scale SDN network, this paper optimizes the controller placement from improving resilience and reliability, puts forward Two-stage Controller Placement (TSCP) method: uses the similarity of node to divide control domain, enhances the connectivity among device in the control domain; selects controller set with the minimum average rate of control path loss as the controller placement for improving the reliability. By constraining the size of control domain and the propagation delay among equipment (switches or controller), makes the number of switch in control domain equalization and the controller placement reasonable. By defining performance indexes, compare GCP algorithm and K*-means algorithm with TSCP algorithm in experiments, the results show that TSCP algorithm can optimize the scale of control domain, balance the number of switches in control domain, reduce the number of controllers, and perform better in network elasticity and reliability.

    参考文献
    相似文献
    引证文献
引用本文
分享
文章指标
  • 点击次数:
  • 下载次数:
  • HTML阅读次数:
  • 引用次数:
历史
  • 收稿日期:2020-01-03
  • 最后修改日期:2020-01-12
  • 录用日期:2020-03-18
  • 在线发布日期:
  • 出版日期: