Sparse Recovery for DOA Estimation With a Reflection Path

In a low-angle tracking scenario, it is difficult to get an accurate estimation of the direction-of-arrival (DOA) due to the presence of reflection path. Direct and reflected signals are highly correlated, which degrades the performance of conventional DOA estimation methods. To solve this problem, we develop a new method for DOA estimation by leveraging the l(2,1)-norm minimization technique in a sparse recovery theory. To be more specific, we treat the direct signal and the corresponding reflected signal as a group, where we encourage group sparsity by applying l(2,1)-norm minimization. We use the information of reflected signal to overcome the effect of reflection path. Moreover, we also propose another sparse recovery-based method which uses information of the predetermined reflection path coefficient under some certain conditions. A method of choosing the regularization parameter is proposed to ensure the robust sparse recovery. Meanwhile, the Cramer-Rao bound (CRB) of low-angle signals is also derived as there is no direct derivation of it. Numerical results demonstrate that the proposed methods can yield superior performance in solving the DOA estimation than the existing methods.