The effect of coherence on the detection performance of quadratic array processors is examined using an exponential‐power‐law model for the signal wave‐front correlation in an uncorrelated noise field. Detection performance is quantified by the small‐signal deflection criterion, which is first reviewed in the general setting of quadratic detectors, and then applied to matched‐field and optimal quadratic beamformers. It is shown that the detection performance of matched‐field beamforming is substantially degraded for large arrays and typical coherence lengths. The optimal quadratic processor reduces this coherence loss, however, at a cost of greatly increased processor complexity. Several suboptimal beamformers with reduced computational load are also developed, and their detection performance is compared to that of the matched‐field and optimal beamformers. One suboptimal processor, the subarray beamformer, has proved outstanding for this application in three respects: It can realize within 1 dB of optimal performance, is robust over a class of correlation functions, and entails a computational burden no greater than full‐array matched‐field beamforming.

This content is only available via PDF.
You do not currently have access to this content.