Atomic Norm Regularized Tensor Completion for Angle Estimation in Bistatic MIMO Radar With Sensor Failures

Faulty array sensors pose a significant challenge in accurately determining the target angles in bistatic multiple-input multiple-output (MIMO) radar systems. To remedy this, we propose a novel method utilizing atomic norm regularized tensor completion for joint estimates of the direction of departure (DOD) and direction of arrival (DOA). We first develop a third-order covariance tensor model for the sensor array data in bistatic MIMO radar and cast the reconstruction of missing entries caused by faulty array sensors as a low-rank tensor completion (LRTC) problem with structurally missing entries. By leveraging the Vandermonde structure inherent in the factor matrices from the CANDECOMP/PARAFAC (CP) decomposition of the covariance tensor, we impose atomic norm constraints on the factor vectors to formulate an atomic norm regularized tensor completion model. Then, an efficient optimization algorithm based on the alternating direction method of multipliers (ADMM) is developed to solve the proposed model. Moreover, to enforce the Vandermonde structure of the factor matrices, we integrate difference coarray processing to rectify the factor matrices during iterations. Finally, joint estimates of DOD and DOA are obtained by applying the estimation of signal parameters via rotational invariance techniques (ESPRIT) algorithm to the Toeplitz matrices derived from the optimized vectors. Extensive simulations demonstrate that the proposed algorithm achieves superior angle estimation accuracy and computational efficiency compared to state-of-the-art algorithms when handling array sensor failures.