Adaptive Robust Radar Target Detector Based on Gradient Test

The exact knowledge of the signal steering vector is not always known, which may result in detection performance degradation when a mismatch occurs. In this paper, we discuss problem designing robust radar target detector background Gaussian noise whose covariance matrix unknown. To improve robustness to mismatched signals, random perturbation that follows complex normal distribution added under alternative hypothesis. Since traditional detectors divide parameters into real parts and imaginary are sometimes difficult obtain, new robust, parameter gradient test derived directly from data. Moreover, CFAR property proven. assessment indicates exhibits suitable signals.