Acoustic Velocity-Independent DOA Estimation for L-Shaped Array via Modified Decoupled Atomic Norm Minimization

In this paper, a DOA estimation method for underwater application is proposed. In the first stage, a semi-positive definite programming (SDP) model on the L-shaped array mutual covariance matrix (MCM) is constructed with the principle of decoupled atomic norm minimization (DANM). This model allows a more accurate reconstruction of the MCM for the L-shaped array. In the second stage, an omnidirectional velocity independence (VI) method is proposed to eliminate the acoustic velocity factor. Numerical simulations show that the proposed method not only achieves effective DOA estimation in the environment of the unknown acoustic velocity but also its performance, especially under high SNR conditions, is significantly improved.