Performance of statistical inference methods for the energy estimation of multiple sources

This article considers statistical inference of the transmit powers of multiple signal sources by a sensor network. Using random matrix methods, an exact expression of the posterior probability of the joint transmit powers is derived. This expression is used to implement the associated ML and MMSE detectors of the joint powers. These are compared for small system sizes against an asymptotically unbiased estimator obtained from large dimensional random matrix theory.