I'm trying to learn about the SAR image formation process. From reference such as Carrara's Spotlight Synthetic Aperture Radar book, it appears that after various compensations, a direct 2D fft can be applied to 'compress' the data.
I understand range is in 'slow time', and crossrange/azimuth is in 'fast time'. However, wouldn't applying the 2D fft directly result in the frequency domain? What are the units in range/crossrange for raw, unfocused SAR data that allow us to directly apply a 2D fft to get a 'compressed' image (I believe this would result in a single look complex image?).