Statistical inference of internal network loss and topology
R. Cáceres, Nick G. Duffield, Joseph Horowitz, Francesco Lo Presti, Don Towsley · ACM SIGMETRICS Performance Evaluation Review · 1999
The use of inference from end-to-end multicast measurements has recently been proposed to find the internal characteristics in a network. Here we describe statistically rigorous methods for inferring link loss rates, and their application to identifying the underlying multicast topology.