【24h】

Modeling TCP Loss Recovery Latency with the Number of Retransmissions

机译:使用重传次数对TCP丢失恢复延迟进行建模

获取原文
获取原文并翻译 | 示例

摘要

So far, several analytic models have been proposed to predict the performance of transmission control protocol (TCP) such as steady-state throughput. However, for more detailed performance analysis of various TCP implementations, the fast recovery latency during which packet losses are retransmitted should be considered based on the relevant strategy. In this paper, we derive the loss recovery latency of three TCP implementations including TCP Reno, NewReno, and when selective acknowledgement (SACK) option is used. Specifically, the number of lost packets each TCP sender detects and retransmits during fast recovery is considered. Thereby, the proposed model can differentiate the loss recovery latency of TCP using SACK option from TCP NewReno. By numerical results verified by simulations, we evaluate that the proposed model can capture the precise latency of TCP loss recovery period.
机译:到目前为止,已经提出了几种分析模型来预测传输控制协议(TCP)的性能,例如稳态吞吐量。但是,为了对各种TCP实现进行更详细的性能分析,应根据相关策略考虑重发丢包的快速恢复延迟。在本文中,我们推导了包括TCP Reno,NewReno和使用选择性确认(SACK)选项在内的三种TCP实现的丢失恢复延迟。具体来说,考虑每个TCP发送方在快速恢复期间检测到并重新传输的丢失数据包的数量。因此,提出的模型可以区分使用SACK选项的TCP的丢失恢复延迟与TCP NewReno。通过仿真验证的数值结果,我们评估了所提出的模型可以捕获TCP丢失恢复期的精确延迟。

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