Tensor-Based Robust Adaptive Beamforming for Multiple-Input Multiple-Output Radar under Random Mismatch Scenarios

Adaptive beamforming for multiple-input multiple-output (MIMO) radar systems suffers from performance deterioration under the scenarios with multiple random mismatches. This paper explores the theory of tensor algebra and exploits the inherent multidimensional structure of data matrix received by MIMO radar systems. For dealing with the beamforming problem induced by multiple random mismatches including steering vector error, mutual coupling, sensor position error, and coherent local scattering, we develop a robust method based on a third-order tensor in conjunction with a gradient-based optimization process. The proposed method captures the multidimensional structure information embedded in the data matrix received by a MIMO radar and produces appropriate estimates for transmit and receive signal direction vectors required for beamforming. Using a third-order tensor helps to alleviate the effect of the multiple random mismatches in the tensor data domain. The gradient-based optimization process further enhances the capabilities of the third-order tensor in estimating transmit and receive signal direction vectors for adaptive beamforming of a MIMO radar. The main computational complexity of the proposed method is dominated by a trilinear alternating least squares algorithm and the well-known gradient-based algorithm. The proposed method provides better performance than the existing robust methods. Simulation results are presented to confirm the effectiveness of the proposed method.