Underwater passive detection and spatial coherence testing

A problem in underwater passive listening is to detect a possible noise source in a given direction when background noise and other sources are present. It is assumed that the signal to be detected is perfectly coherent spatially. If the correlation matrix of the outputs of the sensors without the signal is known, it is possible to compute both an estimation and a detection test for the signal. When this matrix is unknown, which is the case in passive listening, an estimation of the signal can be obtained by an adaptive antenna thanks to the perfect coherence hypothesis. But in order to find a detection test, the shape of the spatial coherence matrix of the background noise must be known. Then the spectral density of this noise can be estimated even in the presence of the signal and of other coherent sources. A detection test using this estimation and the output of the adaptive antenna is proposed. Theoretical asymptotic results and simulations are presented.