Adaptive array processing for distributed sources

We consider the problem of detecting a spatially distributed source by an array of sensors. Source parameters such as signal power, direction, and angular spread, as well as the noise power, are unknown to the detector. Using the generalized likelihood ratio (GLR) approach we derive the structure of the detector under different assumptions on which parameters are known or unknown. The performance of the detectors is evaluated and the effect of angular spread is investigated. It is shown that as angular spread increases, the GLR detector provides significant advantages over the conventional beamformer based detector.