Joint 2-D DOA estimation using gridless sparse method

A novel gridless sparse method (GSM) is proposed to estimate two-dimensional (2-D) direction-of-arrival (DOA) using L-shaped arrays. The angular space is transformed to a frequency one and a new model is constructed in the frequency field. Based on the new model, the covariance matrix is reparameterized by a positive semidefinite Toeplitz matrix. Covariance fitting criterion and semidefinite programming are used to estimate DOAs in the continuous range. Compared with traditional 2-D DOA estimation methods, there is no need to discretize the whole angular space which can cause modeling error and increase computation loads. Moreover, the proposed method can get pair-matching automatically.