Structured covariance estimation for airborne radar STAP via two nuclear norms

This paper presents the special structure of the clutter covariance matrix (CCM) of airborne radar systems. In general, the CCM and the permutated CCM by the definition are both modeled as a summation of low-rank Kronecker products in the space-time adaptive processing (STAP). To handle original and permutated CCM structures, a novel optimization estimation function is founded with two nuclear norms. Resorting to the proximal gradient algorithm, we propose a new method to obtain a unique solution for the structured CCM. In the end of the paper, we assess the prominent performance of the proposed method supported by simulated and measured data of the airborne radar STAP.