计算机工程与应用 ›› 2009, Vol. 45 ›› Issue (3): 97-100.DOI: 10.3778/j.issn.1002-8331.2009.03.028

• 网络、通信、安全 • 上一篇    下一篇

多逻辑链路中反馈信息共享令牌流量调度算法

梁 根1,郭小雪2,秦 勇1,3,魏文红3   

  1. 1.茂名学院 教育信息技术中心,广东 茂名 525000
    2.茂名学院 理学院,广东 茂名 525000
    3.华南理工大学 计算机科学与工程学院,广州 510641
  • 收稿日期:2008-07-18 修回日期:2008-10-20 出版日期:2009-01-21 发布日期:2009-01-21
  • 通讯作者: 梁 根

Sharing token traffic scheduling algorithm in multi logical links based on feedback information

LIANG Gen1,GUO Xiao-xue2,QIN Yong1,3,WEI Wen-hong3   

  1. 1.Education Information and Technology Center,Maoming University,Maoming,Guangdong 525000,China
    2.College of Science,Maoming University,Maoming,Guangdong 525000,China
    3.Department of Computer Science & Engineering,South China University of Technology,Guangzhou 510641,China
  • Received:2008-07-18 Revised:2008-10-20 Online:2009-01-21 Published:2009-01-21
  • Contact: LIANG Gen

摘要: 流量调度的目标是提高网络资源的利用率,改善网络应用的性能,目前,国内外已经有大量的流量调度算法研究工作,但是这些算法不能很好地适应宽带接入中多逻辑链路的动态性、实时性、分布性等特点。针对多逻辑链路环境下流量调度机制存在的问题,分析了链路间流量调度特点,设计一种共享令牌缓冲池流量调度模型,提出一种基于反馈信息的共享令牌流量调度算法,该算法根据令牌缓冲池中可用令牌的资源信息对多逻辑链路流量进行反馈调度。实验结果表明,该算法可以有效地减少链路访问延时,而且链路流量的平滑性、吞吐率等都比其他传统调度算法要好。

关键词: 流量调度, 令牌缓冲, 反馈, 负载均衡

Abstract: Traffic scheduling which aims at improving network resource utilization and network application performance.Currently,much research can be found about traffic scheduling algorithm.However,since logical links in broadband network access are dynamic,real-time and distributed,those scheduling algorithms did not fit for the case well.In this paper,the feature of traffic scheduling in multi logical links is discussed,a model of sharing token buffer traffic scheduling is also designed.A sharing token traffic scheduling algorithm in multi logical links based on feedback information is presented.In this algorithm,the traffic scheduling of links which is based on the feedback information of token numbers in sharing token buffer.Experimental results demonstrate that this algorithm diminish latency,the shape and throughout capacity of traffic is better than other traditional algorithms.

Key words: traffic scheduling, token buffer, feedback, load balancing