Estimate Link-Level Transition Matrix from End-to-End Observations

Network tomography aims to estimate network characteristics from end-to-end (e2e) observation, e.g. link-level loss rate and delay distribution. In this paper, a method is proposed to estimate link-level transition matrix from e2e observations. The estimate obtained from the proposed method is an MLE. Simulation study shows the estimate obtained by the proposed method is very close to the direct measurement that makes it possible to use the estimated characteristic to predict the performance of a link or a path for a short period.