Clutter scattering function estimation and ground moving target detection from multiple STAP datacubes

Methods for estimating the clutter scattering function and detecting ground moving targets, using pulse-Doppler surveillance radar data, are described. The imaging problem is cast as one of structured covariance estimation with time-varying measurement models and illumination patterns. An expectation-maximization (EM) algorithm is derived and the computational issues arising from its use are discussed. The detection algorithm uses the estimated clutter statistics and a time-varying target model in a standard generalized likelihood ratio test (GLRT).