Abstract:

A new router-aided multicast network tomography method is proposed to estimate the spatial-temporal distribution of network performance. Based on MINC (Multicast-based Inference of Network-internal Characteristics), statistical results of background data flows measured by routers are used to improve the correctness of multicast probe even when the probe speed is low. With combination of the maximum likelihood estimation and the aided measurement results, the estimation error is reduced significantly, and the worse influence on background data flows caused by the increase of probe speed is avoided. Simulation shows the validity of the method.

Key words: network tomography, multicast probe, router-aided