Iterative MMSE Algorithm of ISAR Image Reconstruction

The focus of the present work is on the Inverse Synthetic Aperture Radar (ISAR) image reconstruction procedure based on iterative minimization of mean square errors (MMSE) in the estimation of the object's invariant geometric parameters. The ISAR geometry and kinematics are analytically described in two-dimensional (2-D) coordinates. The vector equation for estimation of the invariant geometric parameters and the matrix equation for calculation of errors in the estimates are presented. A linear frequency modulation (LFM) ISAR signal from scattering points located at the nodes of a uniform grid during inverse aperture synthesis is used as an approximation matrix function. In contrast to Fourier transform image reconstruction, the azimuth resolution properties of the proposed method do not depend on the number of measurements, i.e. the synthetic aperture length. The number of the measurements is defined by the number of the evaluated geometric parameters, the object's scattering points, which is the main advantage of the proposed MMSE method, and the main contribution of the present study. To prove the validity and correctness of the developed iterative MMSE algorithm, numerical experiments are performed. The computational results demonstrate high-resolution images, unambiguous and convergent estimates of the scattering point intensities of the object from limited simulated ISAR data.