A method of root-sparse asymptotic minimum variance off-grid DOA estimation based on unitary transformation

To enhance the accuracy and efficiency of direction of arrival (DOA) estimation, an improved method of root-sparse asymptotic minimum variance off-grid DOA estimation is proposed in this paper. By utilizing unitary transformation, the signal of covariance matrix in complex domain is transformed into real domain, which reduces computation and runtime of iteration. Moreover, the unitary transformation has the function of bidirectional smoothing, which improves estimation performance and angle resolution of coherent sources. Additionally, combined with the off-grid model, grid points' positions are directly used as estimation parameters for iterative calculation. It overcomes modeling errors caused by off-grid gaps and improves azimuth estimation accuracy, particularly in scenarios with small snapshots and low signal to noise ratio (SNR) . Simulation results and lake experiment data demonstrate the effectiveness of the proposed method, which will be highly valuable for engineering application.