【24h】

Evaluation of Packet Transmission Delay Variation in the G/G/1 System

机译:G / G / 1系统中数据包传输延迟变化的评估

获取原文

摘要

Analyzing the performance parameters of IP-networks when processing multimedia streams is a very important task. There are many approaches to evaluating the quality of service parameters in the G/G/1 system. Changing the packet delay in the network is a very significant parameter that determines the quality of traffic processing. It is particularly important for multimedia streams. The delay variation is generally defined as a packet jitter. However, the analysis of the delay variation is often based on assumptions that do not allow the parameters to be determined with the required accuracy. This paper presents a new approach to defining packet delay variation in the G/G/1 system as delay variation. The presented approach is based on approximation of arbitrary distributions by hyperexponential distributions, i.e. modeling the G/G/1 system by the H_2/H_2/1 system. The EM algorithm is used to estimate the parameters of hyperexponential distributions. The paper presents the results of simulation. The packet delay variation was evaluated when processing traffic registered on a real network, CBR traffic, traffic with Pareto distribution of time intervals between packets and packet lengths, and traffic with exponential distribution of time intervals between incoming packets. Due to the fact that CBR traffic has explicit correlated properties, it can be noted that the presence of correlation inherent in CBR traffic leads to a decrease in delay variation.
机译:处理多媒体流时的IP网络性能参数是一个非常重要的任务。在G / G / 1系统中评估服务质量参数有许多方法。更改网络中的数据包延迟是一个非常重要的参数,用于确定流量处理的质量。它对多媒体流尤为重要。延迟变化通常被定义为分组抖动。然而,延迟变化的分析通常基于不允许以所需的精度确定参数的假设。本文提出了一种新方法,可以将G / G / 1系统中的数据包延迟变化定义为延迟变化。所提出的方法基于过度开放分布的任意分布的近似值,即通过H_2 / H_2 / 1系统建模G / G / 1系统。 EM算法用于估计过度开发分布的参数。本文提出了模拟结果。当在真实网络上登记的流量,CBR流量,CBR流量,数据包和分组长度之间的时间间隔的流量分发时,评估分组延迟变化,以及传入分组之间的时间间隔的指数分布的流量。由于CBR流量具有明显相关性的性质,可以注意到,CBR业务中固有的相关性导致延迟变化的降低。

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