Performance analysis of mis-modeled estimation procedures for a distributed source of non constant modulus

The problem of estimating the bearing of a single, far-field, non constant modulus source, surrounded by local scatterers using passive sensor array measurements is addressed. An associated source bearing estimation problem is formulated, and the Cramér-Rao lower bound is evaluated. Estimation procedures that assume the source is of constant modulus and therefore suffer from mis-modeling errors are investigated. Specifically, the performance of the application of the maximum likelihood (ML) estimator designed for a constant modulus source performing on a non constant modulus source is evaluated, and the exact degradation in performance is quantified as a function of the source's empirical variance.