Journal of Computer Applications ›› 2014, Vol. 34 ›› Issue (3): 632-634.DOI: 10.11772/j.issn.1001-9081.2014.03.0632

• Network and communications • Previous Articles     Next Articles

Active queue management algorithm of queue delay control

WU Dong   

  1. School of Information Science and Technology, Zhanjiang Normal College, Zhanjiang Guangdong 524048, China
  • Received:2013-08-19 Revised:2013-10-21 Online:2014-03-01 Published:2014-04-01
  • Contact: WU Dong

控制时延的主动队列管理算法

吴东   

  1. 湛江师范学院 信息科学与技术学院,广东 湛江524048
  • 通讯作者: 吴东
  • 作者简介:吴东(1981-),男,广西合浦人,讲师,硕士,CCF会员,主要研究方向:网络拥塞控制。
  • 基金资助:

    广东省自然科学基金资助项目;湛江市科技攻关计划项目

Abstract:

To solve the problem that the queue delay cannot meet the demand of media applications, such as VoIP, real time video and remote video conference in the existing Active Queue Management (AQM) algorithms, a new AQM algorithm named DCQA (Direct Control Queue Delay Algorithm) was proposed. In this new algorithm, to control the queue delay below the expected value, identifying the packet dropping and other corresponding processes were done before a new packet entered the router buffer. The packet dropping rate was computed by PID controller. Then, a simulation in three network environments was conducted to compare the algorithm performance of DCQA and CoDel algorithm. The experimental results show that the queue delay can be effectively controlled and high utilization of link is obtained in DCQA. The detailed data of utilization of link are 99.93%, 99.88% and 99.95%. At the same time, the simulation results show that the probabilities of the queue delay below the expected value are 50.45%, 51.59% and 52.4%, improved by 3.6%, 40.53% and 50.69% compared with the CoDel algorithm. Therefore, DCAQ is applicable for streaming media transmission.

Key words: Active Queue Management (AQM), buffer, queue, Quality of Service (QoS), delay

摘要:

针对现有的主动队列管理(AQM)算法造成的队列时延无法满足VoIP、音视频等流媒体传输需求的问题,提出一种直接控制队列时延的主动队列管理算法——DCQA。该算法使用PID控制器计算路由器缓存的数据包丢弃概率,用其对即将进入缓存排队的数据包做丢包判断并采取相应动作,以控制队列时延在期望值以下。实验仿真了3种网络环境下DCQA的性能,链路利用率分别是99.93%、99.88%和99.95%。并且,队列时延分别有50.45%、51.59%、52.4%被控制在期望值以下,比CoDel算法分别提高了3.6%、40.53%、50.69%。实验结果表明,DCQA在不同的网络环境中都可以获得较高的链路利用率,而且控制队列时延的能力优于CoDel算法,适用于流媒体的传输。

关键词: 主动队列管理, 缓存, 队列, 服务质量, 时延

CLC Number: