Sonar array signal processing for sparse linear arrays

Acoustic signals which propagate through the ocean have wavefronts which can differ significantly from the “planar wavefronts” assumed in array signal processing. In this paper we investigate the performance of the the minimum variance distortionless response (MVDR) beamformer and the previously introduced Fourier integral method (FIM), when applied to real sonar data from a sparse linear array. It is shown that as long as the array is calibrated, MVDR will perform very well, but when sensor gain/phase errors are present the performance of the MVDR beamformer is degraded. FIM however seems to always perform very well, and a modified version of FIM is shown to be at worst comparable to a well calibrated MVDR beamformer