Adaptive beamforming against reverberation for a three-sensor array

Abstract

Two algorithms for beamforming and bearing estimation of echoes from an active sonar transmission in a strongly reverberant bistatic environment are described. The receiving array consists of a single omnidirectional sensor and two collocated orthogonal dipole sensors, and is deployed in a bistatic configuration. Both beamforming algorithms are based on minimum-variance techniques. The first algorithm matches the pattern of the minimum-variance beamformed data with that expected from an impulse at a known bearing. The second algorithm reforms the sensor data from the minimum-variance beam response. Fixed-coefficient limaçon beamforming is then applied to estimate the beam power map with reduced reverberation. The detection performance of both techniques is evaluated by injection of a synthetic target echo into experimental reverberation data. The results suggest an enhanced array gain against reverberation of the order of 3 dB for reasonable values of signal strength and probability of false alarm, compared to a direct application of fixed-coefficient limaçon beamforming. The root-mean-squared bearing error for both techniques is reduced significantly, when compared to the limaçon beamformer, by factors varying from 2 to 5.