Modified Direction-of-arrival Estimation Approach of Virtual Antenna Array Based on Fourth Order Statistics

Fourth order statistics based multiple signal classification (FO-MUSIC) is a direction-of-arrival (DOA) estimation approach of virtual antenna array. The second order covariance matrix is replaced by the fourth order covariance matrix and virtual array elements are formed, by which the aperture of the antenna array is extended and the resolution of DOA estimation is improved. In order to further improve the resolution of the FO-MUSIC approach without increasing the computational complexity, a modified DOA estimation approach of virtual antenna array based on fourth order statistics is proposed. The proposed approach is named modified FO-MUSIC (MFO-MUSIC) and is dependent on reconstructing the steering vector and the fourth order covariance matrix. The reference position of array is re-localized one elements space away from the uniform linear array, rather than at one end side of the array. The phase difference of the reference array element is no longer zero, so the conjugate data in the fourth order covariance that calculated through the steering vector of the reference array element will not loss. Simulation results show that the resolution of the MFO-MUSIC approach is higher than that of the FO-MUSIC approach.