Parametric GLRT-based Adaptive Target Detection for FDA-MIMO Radar in Gaussian Clutter

This paper addresses moving target detection with limited test and/or training samples for frequency diverse array multiple-input multiple-output (FDA-MIMO) radar under Gaussian environments. To overcome the performance degradation caused by limited test/training samples, the disturbance is modeled as an autoregressive (AR) process, and a parametric one-step generalized likelihood ratio test (OGLRT) detector is developed. As the associated maximum likelihood (ML) estimation is nonlinear and computationally demanding, two asymptotic maximum likelihood (AML) approaches are introduced to substantially reduce computational complexity, which are asymptotically equivalent to ML. The asymptotic detection and false alarm probabilities are derived in closed form, and numerical simulations validate that the proposed scheme greatly relaxes the requirement of the number of the test/training data and significantly improves detection performance compared to conventional methods.