Abstract: As ground target modeling is based on DEM（Digital Elevation Model） with complex targets and is not conducive to image feature observation of typical target in GNSS-R SAR imaging simulation, Monte Carlo method is adopted to model random surface and Euclidean principle is applied to model buildings and trees, which builds flexible and controllable imaging scenes. The navigation satellites are used as signal source, and the airborne receiver is served as a signal receiving platform, and the GNSS-R signal reflected by the imaging scene is constructed as an echo signal, at last, BP algorithm performs imaging simulation as imaging processing method. Simulation results show that the image generated in this simulation scenario can reflect the structural characteristics and differences in scattering intensity of different ground targets. Prove that the constructed models can be applied to the verification of GNSS-R SAR imaging simulation system.

Key words: Global Navigation Satellite System（GNSS）, Synthetic Aperture Radar（SAR）, ground target modeling, navigation satellites, BP algorithm

摘要： 针对GNSS-R SAR对地成像仿真中，由于地面目标建模以目标复杂的实地勘测DEM（Digital Elevation Model）为基础而导致的典型目标影像特征不利于单独观察的问题，基于蒙特卡罗方法对随机地表建模，基于欧几里德原理对建筑和树木建模，构建结构灵活可控的成像场景。利用导航卫星作为辐射源，机载接收机作为信号接收平台，成像场景反射的GNSS-R信号作为回波信号，BP算法作为成像处理算法，进行成像仿真。仿真结果表明，该仿真场景下生成的图像可反映不同地面目标的结构特点和散射强度差异，证明了所构建的地面目标模型能够应用于GNSS-R SAR成像系统的仿真验证。

关键词: 全球导航卫星系统（GNSS）, 合成孔径雷达（SAR）, 地面目标建模, 导航卫星, BP算法