A Method Of Detecting Ground Micro-motion Target In Multichannel Airborne SAR

In recent years, the research on synthetic aperture radar ground moving target detection(SAR-GMTI) is mainly focused on the target with constant velocity or uniform acceleration and second-order polynomials approximation is used to refocus imaging. But indeed, the moving target or any structure on it may undergoes micro-motion dynamics, such as mechanical rocking, vibrations or rotations. Specifically, the traditional aforementioned range model is not suitable for describing the range from a radar to the micro-motion targets.This thesis firstly establish the geometrical models of three kinds of micro-motion targets in multichannel SAR :pelagic scattering point with rocking, ground scattering point with vibration and rotation.The application conditions and limits of traditional polynomial expansion theory are compared with that of paired echo principle,thus obtaining that if the periodicity motion is more accurate to use paired echo principle to expand the range.Secondly, this thesis introduces a dual channel processing method to detect the micro-motion target and suppress clutter in data field and analyzes the limitation and inaccuracy of this method. On the basis of ghost points paired echo principle and SAR-RD algorithm, we derive the interferometric phase of the echo among channels, conclude the vector expression in image domain,and analyze the azimuth defocusing in single-channel caused by micro-motion. For exactly describing the characteristics of micro-Doppler, two normalization, the amplitude of the echo of different ghost points and the correlation coefficient of the spatial steering vectors between each ghost point and clutter, are introduced. The result of simulation demonstrates that making use of the feature of ghost points to detect the ground moving target can obtain a good improvement factor in strong clutter background.