Adaptive spatial matrix filter design with application to DOA estimation

The performance of passive acoustic signal processing techniques can become severely degraded when the acoustic source of interest is obscured by strong interferences. An effective approach to the design of adaptive matrix filters was proposed. The idea is to minimize the output power of the transformed data subject to a set of nonlinear inequality constraints imposed on error between the desired response and the designed filter in passband and stopband. These inequality constraints ensure that the stopband attenuation and the passband deviation satisfy the prescribed specifications, thereby, guarantee that signals of interest distortionlessly pass the filter while those of no interest are cancelled effectively. Computationally efficient convex formulations for the adaptive matrix filter design problems are derived using second-order cone programming (SOC), which can be solved by standard interior point method. Computer simulations show that the proposed method can produce nulls in directions of interferences, thus effectively suppresses strong interferences in stopband. As a result, the ability of the DOA estimator in the following procedure is greatly enhanced.