Joint DOA and Doppler frequency estimation for MIMO Radars in the Presence of Array Model Imperfections.

In this paper, the problem of joint direction of arrival (DOA) and Doppler frequency estimation with array model defects is investigated. To this end, a two-dimensional (2D) robust multiple signal classification (MUSIC) algorithm is proposed, which is a generalization of the conventional MUSIC algorithm. The introduced algorithm is based on a colocated multiple-input-multiple-output (MIMO) radar composed of uniform linear arrays, in which a mix of calibrated and uncalibrated antennas exist in its transmitter, while all the receiver antennas are assumed to be uncalibrated. To be specific, the 2D R-MUSIC algorithm adopts a separation scheme to automatically obtain the signals associated with the calibrated subset of transmitter antennas and then eliminates the effect of the signals belonging to the uncalibrated transmitting and receiving antennas. Afterward, the localization is performed by modifying the conventional algorithms in accordance with the separation scheme. In addition, a 2D Genie maximum likelihood (ML) algorithm under the assumption of precisely known errors at the receiver is derived to provide a lower bound benchmark for the introduced method. The simulation results demonstrate the superiority of the proposed algorithms relative to the conventional ones.