Bearing estimation for a distributed source via the conventional beamformer

This paper considers the performance of the conventional beamformer (BF) for estimating the bearing of a single, far-field, narrowband, spatially distributed source with an array of passive sensors. The BF offers significant computational advantage compared to the (asymptotically optimum) maximum likelihood estimators (MLEs) which require searches over some multi-dimensional parameter space. However, the price to be paid for such a reduction in computational complexity is a degradation in performance compared to that of the MLEs. This paper studies the nature of this degradation for “coherently” and “incoherently” distributed sources as a function of key scenario parameters such as signal-to-noise-ratio (SNR), mean angle, spatial spreading of the source, and number of sensors, both mathematically and by simulations