Journal of Systems Engineering and Electronics ›› 2009, Vol. 31 ›› Issue (8): 1930-1934.

• 制导、导航与控制 • 上一篇    下一篇

基于H控制方法的多链路网络拥塞控制

马可芸, 俞立, 张文安   

  1. 浙江工业大学信息工程学院, 浙江, 杭州, 310032
  • 收稿日期:2007-12-20 修回日期:2008-05-08 出版日期:2009-08-20 发布日期:2010-01-03
  • 作者简介:马可芸(1984- ),女,硕士研究生,主要研究方向为时滞系统、网络拥塞控制、神经网络稳定性分析.E-mail:makeyun2@tom.com
  • 基金资助:
    国家自然科学基金(60834003);国家杰出青年科学基金(60525304)资助课题

H control approach to congestion control for multi-link network

MA Ke-yun, YU Li, ZHANG Wen-an   

  1. Coll. of Information Engineering, Zhejiang Univ. of Technology, Hangzhou 310032, China
  • Received:2007-12-20 Revised:2008-05-08 Online:2009-08-20 Published:2010-01-03

摘要: 研究了一类基于主动队列管理(active queue management,AQM)路由器的多链路网络拥塞控制问题.用一个状态空间模型来描述多链路网络流量的动态特性,并将实际链路带宽相对于期望带宽的偏差作为干扰信号,从而用H控制方法来解决网络拥塞问题.在链路带宽变化的情况下,通过丢包率来控制实际窗口大小及队列长度,使得实际窗口大小和队列长度达到期望值,从而达到降低网络拥塞的目的.最后通过两个仿真示例验证了方法的有效性.

Abstract: The congestion control problem is investigated based on a class of multi-link networks with active queue management(AQM) routers.A state-space model is used to describe the dynamic properties of the multi-link network flow,and the discrepancy between available link bandwidth and nominal link bandwidth is considered as a disturbance.Then,the network congestion problem is solved by using the H control approach.The size of actual window and length of queue are adjusted to the reference values by the packet-dropout rate under the change of the available link bandwidth,and thus the congestion control purpose is achieved.Finally,two simulations show the effectiveness of the proposed method.

中图分类号: