Estimating the source-destination traffic matrix of a VPN from access-link loads

We address the problem of estimating the source-destination (SD) traffic matrix of a VPN from measurements of aggregated traffic along all access links. We derive two different estimators by formulating the estimation of the SD-traffic matrix in the form of both deterministic and stochastic optimization problems. One estimator yields the estimates from a snapshot of the access link loads; the other yields estimates from a set of measurements of the access link loads by using a statistical method. The proposed estimators have expressions much simpler than those of the standard SD-traffic-matrix estimators; one estimator has an explicit representation that does not involve matrix inversion. The other estimator is given by a simple matrix formula and does not need any prior SD-traffic matrices. Through simulation experiments, we show that these proposals are superior or comparable to the standard SD-traffic-matrix estimator in term of estimation accuracy.