A Single-Snapshot Robust Beamforming for FDA-MIMO Radar Based on AID$^\text{3}$

Due to the range-angle two-dimensional beam control ability of frequency diverse array multiple-input–multiple-output (FDA-MIMO) radar, it has a wide range of application prospects in mainlobe interference suppression. Several robust beamforming methods for FDA-MIMO radar have been proposed in recent years. However, most of these methods require sufficient independent and identically distributed snapshots, which is difficult to satisfy for FDA-MIMO radar, especially in dynamic environments. Therefore, the problem of FDA-MIMO robust beamforming in sample-starved scenarios is considered in this paper, especially for an extreme case where only a single snapshot can be used for beamforming. This work proposes an improved single-snapshot robust beamforming method for FDA-MIMO radar. We eliminate the desired target component by forward-backward spatial smoothing, expand the number of samples, and construct a robust beamforming problem based on the worst-case performance optimal (WCPO) criterion. To reduce the computational complexity, the original problem can be divided into two convex optimization subproblems using the Kronecker product (KP) structure of the steering vector. The optimal weight vector can be obtained by iteratively solving the two convex optimization subproblems. In addition, the Shrinkage method is introduced to alleviate the mismatch of the covariance matrix caused by the lack of snapshots. Numerical examples show that the proposed method has better performance than the existing competitive algorithms in terms of output signal-to-interference-plus-noise ratio (SINR), robustness, and computational complexity.