Journal of South China University of Technology (Natural Science Edition) ›› 2008, Vol. 36 ›› Issue (9): 20-24.

• Computer Science & Technology • Previous Articles     Next Articles

Active Queue Management Algorithm Based on Congestion Forecast Threshold

Lai Jun  Ye Wu  Feng Sui-li   

  1. School of Electronic and Information Engineering, South China University of Technology, Guangzhou 510640, Guangdong, China
  • Received:2007-10-22 Revised:2007-12-04 Online:2008-09-25 Published:2008-09-25
  • Contact: 赖峻(1979-),男,博士,主要从事高速网络拥塞控制及无线Mesh网络的调度策略研究. E-mail:laijun_2005@126.com
  • About author:赖峻(1979-),男,博士,主要从事高速网络拥塞控制及无线Mesh网络的调度策略研究.
  • Supported by:

    广东省自然科学基金资助项目(31391);粤港关键领域重点突破项目(20060104-2)

Abstract:

In order to solve the problem of sharp decline in performance which exists in Drop Tail gateways and to find a novel algorithm that is easy to implement and can adapt parameters according to the variation of network, an active queue management (AQM) algorithm with explicit congestion notification (ECN) named bandwidth-delay product queue (BDPQ) is proposed. In this algorithm, the congestion in the buffer of the bottleneck gateway is monitored according to the current queue length, and the bandwidth-delay product is used as the threshold for congestion detection. If the queue length exceeds the threshold, the dequeuing packets will be marked according to the ECN policy to notify the congestion to the TCP source. Simulated results show that the proposed algorithm is of high link utilization, short mean queue, low loss rate and good fairness.

Key words: gateway, congestion control, active queue management, bandwidth-delay product queue, explicit congestion notification