距离多普勒成像和多散射点定位

距离多普勒雷达成像系统的基本任务是利用雷达回波信号的距离和多普勒信息重构目标反射率的空间分布。本文把用逆合成孔径雷达对运动目标进行距离多普勒成像当作多散射点定位问题来研究,把成像雷达信号处理的两个关键,运动补偿和成像作为最大似然估计问题一起求解。原则上需进行多维搜索,一般情况下,目标上散射点的个数、位置以及目标的轨道运动都先验未知,这种求解的计算量极大。在若干合理的假设成立时,可引入DFT,免去估计散射点个数和位置,大大减少搜索维数,并可利用FFT,从而大大提高计算效率。这样做的代价是最终的成像质量有所下降。当目标轨道运动模型可用二次的距离时间函数描述时,只需进行二维搜索,文中导出了进一步简化的相位补偿和成像算法,并给出了计算机模拟结果。

Range-Doppler Imaging and Multiple Scatter-Point Localization

The fundamental task of a range-Doppler radar imaging system is to reconstruct the spatial distribution of reflectivity of a target from the returned radar signals containing range and Doppler information. In this paper the range-Doppler imaging of a moving target by an inverse synthetic aperture radar is investigated from the viewpoint of multiple scatter-point localization. Both motion compensation and image formation, which are the two critical steps of signal processing in imaging radars, are solved simultaneously as a maximum likelihood estimation problem. In principle, a multidimensional search is required. Generally, the computational load is extremely heavy because the number and the location of scatter-points on the target and the orbital motion of the target are all unknown a priori. When some reasonable assumptions are valid, we may introduce DFT and avoid estimating the number and locations of scatter-points. Thus the searching dimensions may be greatly reduced. The computational efficiency can be further increased by utilizing FFT, The cost for so doing is a lower final imaging quality. When the orbital motion of the target is described by a quadratic range-time function, only two-dimensional search is needed. A further simplified motion compensation and imaging formation algorithm is derived. Computer simulation results are also given.