基于最小均方误差NUFFT的SAR成像算法

SAR Imaging Algorithm Based on Least-squares Error NUFFT's

在合成孔径雷达成像算法中,距离徙动算法(RMA)在大面积成像以及低波段的情况下成像效果良好,尤其对超宽带雷达成像有较好效果。但是RMA需要精确的插值,这就会引起庞大的计算量。文中将非均匀快速傅里叶变换(NUFFT)应用到RMA成像算法中,用NUFFT来替换RMA中的Stolt插值和距离向的逆快速傅里叶变换,去除了Stolt插值而引起的巨大的计算量,仿真结果证明了该算法的有效性。

In SAR imaging, Range Migration Algorithm （RMA） can be implemented effectively under the condition of the big area and low frequency, especially in UWB-SAR imaging. However, RMA must include the so-called stolt interpolation, which will cause the huge amount of calculation. In this paper, NUFFT will be applied in RMA, and our approach consists in substituting both the stoh interpolation and the final range inverse FFT by a single NUFFT, wipe off skillfully vast calculation arised from stolt interpolation. Numerical simulations illustrate the performance of the new method.