论文部分内容阅读
为了能简单准确地估计网络内部链路的丢包率,该文利用网络层析技术提出了一种低复杂度的推测方法LCIA(low complexity inference approach)。LCIA方法通过在网络内部节点部署监测器来减小推测丢包率的复杂度。它主要由两步组成,第一步是在网络内部部署监测器把整棵树分成较小深度的若干棵子树。第二步是估计每棵子树的链路丢包率,用到的估计式是叶子节点观测结果的显式表达式来估计丢包率。定性的分析和实验仿真结果表明LCIA在计算复杂度和估计结果的精确度上,都优于已有的方法。
In order to estimate the packet loss rate of network internal links simply and accurately, a low complexity inference approach (LCIA) is proposed by using network layering technique. The LCIA method reduces the complexity of speculating packet loss rates by deploying monitors at nodes inside the network. It consists primarily of two steps. The first step is to deploy a monitor inside the network to divide the whole tree into smaller subtrees. The second step is to estimate the link loss ratio for each subtree. The estimated formula used is an explicit expression of leaf node observations to estimate the packet loss rate. Qualitative analysis and experimental simulation results show that LCIA outperforms the existing methods in terms of computational complexity and accuracy of estimation results.