Single-Snapshot Time-Domain Direction of Arrival Estimation under Bayesian Group-Sparse Hypothesis and Vector Sensor Antennas

In this work, an optimal single-snapshot, time domain, group-sparse optimal Bayesian DOA estimation method is proposed and tested on a vector sensors antenna system. Exploiting the group-sparse property of the DOA and the Bayesian formulation of the estimation problem, we provide a fast and accurate DOA estimation algorithm. The proposed estimation method can be used for different steering matrix formulations since the optimal standardization matrix is computed directly from the knowledge of the steering matrix and noise covariance matrix Thanks to this, the algorithm does not requires any kind of calibration or human supervision to operate correctly. In the following, we propose the theoretical basis and details about the estimation algorithm and a possible implementation based on FISTA followed by the results of our computer simulations test.