Adaptive Transmit Beam Pattern Design for Compressed Sensing-based Direction of Arrival Estimation

Transmitter beam pattern design has a crucial role in direction of arrival (DoA) estimation for radar/sonar applications. For a good DoA estimation performance, power amplifiers are used in saturation, and pencil beams are transmitted. In a practical operation scenario, two modes are used simultaneously to both track registered targets and scan for new targets. In the first mode, transmitter tracks targets that are known with a prior distribution. The other mode is used to scan all possible angles using narrow beams. Many beams should be transmitted to scan all angles. In this study, we propose an alternative scheme that allows both modes to be used simultaneously. The proposed scheme uses specifically tailored beams exploiting the prior information to both track targets with high localization while finding emerging targets. The proposed design is based on the reciprocity property of antennas to develop an optimization problem incorporating prior target distributions through the weighting of a target matrix. As a result, the proposed design expresses an optimization-based adaptive transmitter beam pattern design for both tracked and emerging targets. We design transmit beams under certain power limits per sensor, and our design is tested for several scenarios such as signal-to-noise ratio (SNR)s and various number of transmit beams. For a fixed number of transmit beams, the proposed design localizes both tracked targets with prior distributions and emerging targets with higher DoA estimation performance compared to classical state-of-the-art techniques.