Robust Beamforming Using Iterative Worst-Case Performance Optimization for MIMO Radar

The output signal to noise plus interference (SINR) of the existing MIMO radar beamforming algorithms suffer from the large joint steering vector mismatch. Aiming at this problem, a method using iterative worst-case performance optimization is proposed in this paper. Combining the MIMO radar transceiver array structure, the joint steering vector mismatch error, especially the direction estimation error, is analyzed in theory and the empirical value of direction estimation mismatch for co-located MIMO radar is given. Based on this, the limitation of the large-uncertainty-set-based algorithms is pointed out. The proposed algorithm uses a smaller set to constraint all errors and iterate over the estimated joint steering vector using the corresponding relationship between the joint steering vector and the weighted vector until meet the designed termination conditions. Simulation experiments show that the proposed empirical value is consistent with the theoretical value, and that the proposed algorithm presents outstanding performance in improving the output SINR and stong robustness for large mismatch case.