【摘要】 在分析现有的网络拓扑推断算法的基础上,提出一种改进的基于丢包率的多播网络拓扑推断算法。结合接收节点的层次信息、汉明距离及节点接收的探测包数量,能够同时推断网络拓扑结构和链路丢包率,并根据链路丢包率的估计值动态地调整拓扑推断的判决门限值,提高了推断的准确性。仿真实验证明,与现有算法相比,该算法具有更好的性能。更多还原

【Abstract】 By analyzing the existing network topology inference algorithms,this paper presents an improved multicast network topology inference algorithm based on packet loss rate.It combines the hop count of receivers,the Hamming distance and the number of received probes and can infer multicast network topology and loss performance simultaneously.It adapts dynamically the value of threshold according to the estimation of link packet loss rates,it improves the accuracy of the inference.Simulation results show that compared with the existing algorithms,this algorithm has better performance.更多还原