Optimal steering vector adaptation for linear filters leading to robust beamforming

In many filter applications the exact steering vector is not know, and thus, robust beamforming methods have to be used. In this contribution, an algorithm which achieves robust beamforming via target tracking is proposed. In contrast to existing approaches, the algorithm works on sparse signals with arbitrary steering vector shapes, and the parameters of the algorithm are adapted in an optimal way. This is achieved by deriving and evaluating the probability of detection and false alarm for general steering vector mismatches. These probabilities are used to adjust the parameters, such that the number of false positive and false negative detections is minimal. Simulations confirm the theoretic results and show that the algorithm performs better than other approaches.