Sparse representation based DOA estimation using a modified nested linear array

Recently, a modified nested linear array (MNLA) with increased degrees of freedom (DOF) has been proposed based on the traditional nested linear array. The MNLA has a hole-free difference coarray and larger virtual array aperture, and hence, can be used to perform direction-of-arrival (DOA) estimation by utilizing spatial smoothing multiple signal classification algorithm. As a matter of fact, owing to the increased number of virtual sensors in the resulting difference coarray, enhanced performance of DOA estimation is achievable if sparse representation is properly taken into account. To this end, in this paper, we further investigate this modified array structure for DOA estimation under the framework of sparse representation. We first use the ideal covariance matrix to analyze the maximal level of detectable sources that can be achieved by the MNLA. Then, the sample covariance matrix is studied to perform DOA estimation with sparse representation. Numerical simulations are carried out to perform DOA estimation and to illustrate the effectiveness and superiority of the proposed algorithm.