Parametric velocity synthetic aperture radar: Signal modeling and optimal methods

Velocity synthetic aperture radar (VSAR) is equipped with a linear array to receive the echoes from a radar illuminating area via multiple channels, each of which can reconstruct a reflectivity image for the same stationary scene. Based on analysis of pixel vector sampled among multi-images, VSAR may effectively suppress the strong ground clutter and improve moving target detection and location. In this paper, different Doppler-distributed properties are derived for the moving target and clutter, respectively. Then, we propose a novel parametric statistical model for VSAR by dividing the pixel vector into three components, namely, target, clutter, and noise. Furthermore, a method of adaptive implementation of optimal processing (AIOP-VSAR) is presented for moving target detection. It is shown that the optimum detection performance may be obtained via AIOP-VSAR, particularly for the slowly moving target in an inhomogeneous clutter environment. Also, the Cramer-Rao bounds (CRBs) are derived for the estimation of unknown model parameters, as well as the azimuth locations of moving targets, and the maximum-likelihood methods are proposed to reach these CRBs. Based on the proposed target detection and parameter estimation methods, we present a complete parametric flowchart for VSAR. It is demonstrated that the proposed flowchart may effectively mitigate the "azimuth, location ambiguity" of VSAR and has the super-resolution ability to resolve "velocity layover" for multiple targets. Finally, some detailed numerical experiments and scene simulations are provided to show the effectiveness of the proposed methods.