首页> 外文期刊>International journal of systems science >A robust and high-performance queue management controller for large round trip time networks
【24h】

A robust and high-performance queue management controller for large round trip time networks

机译:强大的高性能队列管理控制器,适用于大型往返时间网络

获取原文
获取原文并翻译 | 示例
获取外文期刊封面目录资料

摘要

Congestion management for transmission control protocol is of utmost importance to prevent packet loss within a network. This necessitates strategies for active queue management. The most applied active queue management strategies have their inherent disadvantages which lead to suboptimal performance and even instability in the case of large round trip time and/or external disturbance. This paper presents an internal model control robust queue management scheme with two degrees of freedom in order to restrict the undesired effects of large and small round trip time and parameter variations in the queue management. Conventional approaches such as proportional integral and random early detection procedures lead to unstable behaviour due to large delay. Moreover, internal model control-Smith scheme suffers from large oscillations due to the large round trip time. On the other hand, other schemes such as internal model control-proportional integral and derivative show excessive sluggish performance for small round trip time values. To overcome these shortcomings, we introduce a system entailing two individual controllers for queue management and disturbance rejection, simultaneously. Simulation results based on Matlab/Simulink and also Network Simulator 2 (NS2) demonstrate the effectiveness of the procedure and verify the analytical approach.
机译:传输控制协议的拥塞管理对于防止网络内的数据包丢失至关重要。这需要用于主动队列管理的策略。最常用的主动队列管理策略具有其固有的缺点,在往返时间长和/或外部干扰较大的情况下,这会导致性能欠佳甚至不稳定。本文提出一种具有两个自由度的内部模型控制鲁棒队列管理方案,以限制队列管理中往返行程时间和参数变化的不良影响。诸如比例积分和随机早期检测程序之类的常规方法由于延迟大而导致行为不稳定。而且,由于往返时间长,内部模型控制-史密斯方案遭受较大的振荡。另一方面,其他方案,例如内部模型控制比例积分和导数,对于较小的往返时间值显示出过慢的性能。为了克服这些缺点,我们引入了一个系统,该系统需要两个单独的控制器同时进行队列管理和干扰抑制。基于Matlab / Simulink和Network Simulator 2(NS2)的仿真结果证明了该程序的有效性并验证了分析方法。

著录项

相似文献

  • 外文文献
  • 中文文献
  • 专利
获取原文

客服邮箱:kefu@zhangqiaokeyan.com

京公网安备:11010802029741号 ICP备案号:京ICP备15016152号-6 六维联合信息科技 (北京) 有限公司©版权所有
  • 客服微信

  • 服务号