THE ELEVENTH IEEE SENSOR ARRAY AND MULTICHANNEL SIGNAL PROCESSING WORKSHOP (IEEE SAM2020)

In this paper, a sparsity-inducing DOA estimation method is proposed for strictly noncircular sources with unknown mutual coupling. In the proposed method, the block-sparse recovery model is firstly formulated via parameterizing the steering vector without losing the array aperture. Then, taking the noncircularity of sources into account, a joint sparsity-inducing framework combined with the reweighted norm optimization is constructed to estimate DOA, where the weighted matrix is structured by the noncircular MUSIC-like (NC MUSIC-like) spectrum function to strengthen the sparsity. Finally, DOA estimation can be realized via screening the position of non-zero blocks of the recovered block sparse matrix. Some simulations are implemented to demonstrate that the proposed method shows the effectiveness and superiority with unknown mutual coupling

DOA estimation; mutual coupling; block-sparse recovery; noncircular sources; reweighted l1-norm optimization