Target detection with persymmetric subspace models for steering vector mismatches in MIMO radars

Abstract

In colocated multiple-input multiple-output radar, the problem of detecting a target embedded in Gaussian noise with unknown covariance matrix is examined for the mismatched case where the true transmit and receive steering vectors are not aligned with the nominal ones, respectively. We adopt subspace models to take into account the steering vector mismatches. With exploitation of persymmetry, adaptive detectors without requiring training data are proposed according to the criteria of generalized likelihood ratio test (GLRT), Wald test, and Rao test. The proposed GLRT and Wald test exhibit constant false alarm rate properties against the noise covariance matrix. Interestingly, we find out that the Rao test does not exist for the detection problem considered. Simulation results show that the proposed detectors bear stronger robustness to the steering vector mismatches than their counterparts.