基于标值更新通信量模型的ATM缓冲区信元丢失率分析方法

A Novel Approach to Estimating the Cell Loss Probability of an ATM Buffer Based on Marked Renewal Traffic Model

提出了一种不同于传统分析方法（如ＭＭＰＰ／Ｄ／１）的直接基于输入通信量带宽分布和自相关特性的信元丢失率分析方法，应用标值更新模型近似实际输入通信量的低阶统计特性，给出了一个基于马尔柯夫链极限概率的信元丢失率估计公式。该方法可以估计突发级的信元丢失率，而且具有算法简单，计算量小，数值稳定性好等特点。

A novel approach is proposed to estimate cell loss probability of an ATM buffer. The proposed approach is directly based on the bandwidth distribution and autocorrelation properties of the input traffic,which is different to the traditional analysis,such as MMPP/D/1.Using the marked renewal model to approximate the actual input traffic,we derive a formula for the cell loss probability expressed in terms of the limiting probabilities of an embedded Markov chain.This approach can give a good estimation to the burst level cell loss probability,and its advantages include simplicity,computational efficiency,and numerical stability.