Beamspace Adaptive Beamforming for Hydrodynamic Towed Array Self-Noise Cancellation

.A beamspace adaptive beamformer implementation for the rejection of cable strum self-noise on passive sonar towed arrays is presented. The approach focuses on the implementation of a white noise gain constraint based on the scaled projection technique due to Cox et al. [IEEE Trans. on ASSP, Vol. 35 (10), Oct. 1987]. The objective is to balance the aggressive adaptation necessary for nulling the strong mainlobe interference represented by cable strum against the conservative adaptation required for protection against signal selfnulling associated with steering vector mismatch. Particular attention is paid to the definition of white noise gain as the metric that reflects the level of mainlobe adaptive nulling for an adaptive beamformer. Adaptation control is subsequently performed through the implementation of a constraint on maximum allowable white noise gain at the output of the adaptive processor. The theoretical development underlying the scaled projection based constraint implementation is reviewed. Towed array data results depicting the performance gain of the new ABF algorithm optimized for strum cancellation relative to that of a more conservative baseline ABF algorithm are presented.