Persymmetric GLRT for distributed target detection in compound-Gaussian clutter plus subspace interference

This research tackles the detection challenge for distributed targets in environments with subspace interference and compound-Gaussian clutter. With interference and target signals occupying unknown coordinates within two separate linear subspaces, and clutter characterized by inverse Gamma distributions with an unspecified covariance matrix, the study introduces a detection scheme. Assuming the availability of training data for estimating the covariance matrix, we harness the persymmetry of the clutter covariance matrix to devise a two-step generalized likelihood ratio test capable of efficiently countering interference. Theoretical analyses demonstrate that the suggested detector affords an asymptotically constant false alarm rate against the unknown covariance matrix. Furthermore, Monte Carlo simulations produced numerical findings showing that the suggested detector outperforms current approaches, especially in cases of limited training data.