相位梯度自聚焦的离散域分析与实现

Analysis on Phase Gradient Algorithm in Discrete Domain and Implementation

通常的自聚焦方法基于低阶相位误差模型，只能估计低阶相位误差，所以不适合高分辨力机载合成孔径雷达ＳＡＲ（ＳｙｎｔｈｅｔｉｃＡｐｅｒｔｕｒｅＲａｄａｒ）的要求．本文研究的相位梯度自聚焦ＰＧＡ（ＰｈａｓｅＧｒａｄｉｅｎｔＡｕｔｏｆｏｃｕｓ）是非模型的相位误差估计，因此对任意阶相位误差都有非常好的自聚焦性能．本文在离散时域和频域探讨和综述了ＰＧＡ的有关问题及其实现方法．最后给出了ＰＧＡ算法与一种基于低阶相位误差模型的ｍａｐｄｒｉｆｔ自聚焦算法比较的实验结果．

Usual autofocus techniques are based on low order phase error model, and they can only estimate low order phase error, therefore they can not be fit for high resolution airborne SAR(Synthetic Aperture Radar). The PGA(Phase Gradient Autofocus) presented in this paper is uniquely not phase error model based, so it has a well deserved reputation for arbitrarily high order autofocus. This paper researches on PGA and its implementation in discrete domain, analyzes and summarizes correlative problems. Finally, some experimental results are given to compare PGA with mapdrift autofocus which is based on limited order phase error estimation.