Applications of adaptive signal processing to the reduction of flow-induced noise in small underwater vehicles

We analytically and experimentally investigate the effectiveness of adaptive filtering for reducing flow-induced sonar interference in small underwater vehicles. Noise and signal fields are modeled and applied to an array whose beamformer is designed to maximize signal to noise ratio. The beamformer output is applied to an adaptive processing system with two auxiliary channels, one for a spatial sample, and one for a temporal sample of interference. Noise, signal, and adaptive filter parameters are examined analytically to define design guidelines for the system and assess performance in a variety of realistic signal and noise situations. The hardware implementation emphasizes architectures that meet the performance requirements and are responsive to the physical constraints of small underwater vehicles. A digital high speed, pipelined transversal filter is used to produce the noise cancelling signal, and the computations for error signal and filter weights are performed in parallel by high speed microprocessors.