Adaptive structured detector and performance assessment in training-limited cases

Abstract

Adaptive detection of a radar target is addressed in Gaussian clutter with persymmetric covariance matrix. Resorting to the persymmetry of covariance matrix, a structured detector of generalized likelihood ratio test is discussed. A more efficient version is further given in real domain. The performance assessment conducted by Monte Carlo simulation confirms its constant false alarm rate property and its advantage of detection performance in the case of insufficient number of training data.