Study On The Imaging Algorithm Of Squint Spotlight Synthetic Aperture Radar

Without the restriction of sunlight and weather, synthetic aperture radar (SAR) can perform all-weather and all-time observation. SAR systems have different modes, like the strip, the scan, and the spotlight modes. Different modes have different applications. The advantage of the spotlight mode SAR is that it can perform high azimuth resolution imaging for the illuminated scene. It is implemented by directing the beam of the antenna to the same area during the synthetic aperture time. In this situation, the azimuth resolution of the spotlight SAR does not depends on the width of the beam of antenna. Therefore, the spotlight mode SAR supplies a feasible way to obtain radar images of a 1-m or higher resolution.In this dissertation, we will focus on how to correct the range migration and reduce high pulse repetition frequency (PRF)。Range migration results from the relative motion between the radar and the target, and cannot be avoided in a SAR system. Usually, the linear part of range migration is called range walk, and the quadratic part is called range curve. Most of spotlight SAR processing algorithms were proposed for the broadside mode, and there is no range walk. For the more general mode, the squint mode, range walk dominates gradually with the increasing of the Doppler centroid. In this case, the imaging algorithms with only range curve considered cannot work well.The range migration algorithm (RMA) is a promising spotlight SAR imaging algorithm and the key to the RMA is the Stolt interpolation. For an actual imaging system, due to the large-storage requirement and computation burden, the interpolation is not practical. The frequency scaling algorithm (FSA) is a preferable approximation of the RMA. It avoids the interpolation of the RMA by an accurate approximation. Unfortunately, the original FSA is based on the broadside mode and does not apply to the squint imaging. The approximation phase error caused by the secondary range compression restricts the performance of the FSA processing for the squint mode.In order to reduce high PRF, the long spotlight aperture is usually divided into several sub-apertures to process. The step transform is an effective sub-aperture processing approach. The quality of the final image is limited by two factors. One is that the offset between two adjacent sub-apertures must be a multiple of the frequency interval. The other is that the high order phase error exists in the coarse resolution data.Considering the aspects mentioned above, we have obtained following achievements.First, the wave-number spectrum extension before the Stolt interpolation of the RMA is analyzed to derive the measure of image quality in the squint imaging. This can be used to evaluate results by other imaging algorithms. Moreover line-of-sight interpolation for the RMA is proposed.Second, in order to compensate the phase error from secondary range compression accurately, we derive the optimal nonlinear frequency scaling function. The nonlinear frequency scaling algorithm works very well even for large squint angles and large imaging scenes.Third, the step transform is combined with autofocus to improve the sub-aperture processing. The minimum-entropy filter using an adaptive order polynomial model is designed to estimate and remove the phase error.Finally, the simulation and processing results for the space-borne spotlight mode SAR system and the space-borne sliding spotlight mode SAR system are given. In addition, some practical airborne data are processed by the chirp scaling algorithm, and the results confirm the validity of the spotlight SAR imaging algorithm presented in this paper.