为了提高组播网络中链路丢包测量的正确性,提出了以2-state Gilbert模型为链路丢包模型的算法.该算法利用组播网络各链路的丢包率在空间上具有相关性的特点,基于网络端到端的测量,采用最大似然算法(ML)计算网络中各节点状态,然后根据各节点的状态计算出网络中各链路的丢包率.仿真实验表明,该算法具有准确度高、收敛速度快的特点.
To improve the accuracy of measuring the rate of packet loss in a muhicast network, an improved algorithm with the 2-state Gilbert model is proposed. In a multicast network, the rates of packet loss among linked routers are correlated to some degree. Based on the network end-to-end measurement, ML algorithm is adopted to compute the state of each node in the network and calculate the packet loss rate based on each node state. The simulation results verify that the improved algorithm is capable to measure loss rate more accurately with faster convergence.