Study On Multi-mode SAR Imaging Algorithm And Parameter Estimation Method

Synthetic aperture radar(SAR) is a powerful and well-established microwave remote sensing technique that enables two dimensional(2-D) high-resolution measurements of the Earth’s surface from long distance. SAR also is a powerful and effective tool for the retrieval of geographical information because of its capability to work under all weather conditions, day and night. It has been widely used in the national defence and civil industry. For applications with different requirements, many imaging modes have been developed in the literature, such as, stripmap SAR, spotlight SAR, sliding spotlight SAR, and Terrain Observation by Progressive Scans SAR(TOPS SAR). In addition, due to the flexible beam steering, SAR can work in different configurations, for example, broadside, squinted and forward looking. Among different imaging modes and configurations, the echo signals have their own characteristics. As such, the imaging algorithms are different. This dissertation focuses on multi-mode SAR data processing. Based on the imaging geometry and by the analysis of the echo data, some useful imaging approaches are proposed. The main content of the dissertation is summarized as follows.1. Compared with spaceborne TOPS SAR, the scanning speed is faster and the steering angle is larger in airborne case. Using these characteristics, a novel full aperture imaging algorithm is proposed for airborne TOPS SAR. Firstly, according to the TWO-STEP method, azimuth pro-filtering is carried out to eliminate the aliasing caused by the beam steering, and thus obtain the 2-D spectrum without blurring. Then the Nonlinear Chirp Scaling(NCS) method is used to complete the range cell migration(RCM) correction and range compression. After that, the signal is focused in the Doppler domain by SPECAN technique to avoid aliasing in the focused image. Finally, geometric distortion correction is carried out via a Chirp-Z Transform(CZT). Without interpolation and coordinate transformation, the algorithm is high in computational efficiency. The imaging results from simulation and real data have demonstrated the effectiveness of the method.2. For squinted TOPS SAR mode, the generalized polar formation algorithm(GPFA) is improved and an imaging algorithm, which firstly carries out linear range walk correction(LRWC) and then applies the PFA interpolation, is proposed. The LRWC caneffectively mitigate the range-azimuth coupling effect and simplify RCM correction. Besides, different from the coordinate rotation operation which is commonly used in the PFA for squinted spotlight SAR, after performing LRWC, the azimuth sampling interval is still uniform. This character provides great benefits to the azimuth interpolation. For the azimuth-variation of Doppler rate induced by the antenna beam steering and LRWC, the azimuth NCS(ANCS) is implemented. Therefore, the azimuth focus depth is improved greatly and the scene size which can be well-focused is increased. Both Simulation and real data results show the effectiveness of the proposed algorithm.3. An imaging algorithm based on azimuth resampling is proposed for highly squinted SAR imaging with several modes. Linear range walk yields significant range-azimuth coupling effect in a highly squinted SAR. Although LRWC technique can effectively mitigate such coupling effect, it causes azimuth-variation in the resulting signal and, as such, the so-called “azimuth-shift invariance” property becomes invalid. In order to eliminate the azimuth-variation, a new spectrum processing approach based on azimuth resampling is proposed. After performing LRWC, the azimuth resampling is carried out in the two-dimensional frequency domain and transforms the signal spectrum to be equivalent to that of a broadside SAR. For squinted beam-steering SAR(BS-SAR), e.g., spotlight SAR, sliding spotlight SAR, and TOPS SAR, the azimuth resampling is combined with the azimuth signal reconstruction algorithm. As a result, both the azimuth-variation, induced by the LRWC, and the aliasing, caused by antenna beam steering, can be avoided. Therefore, after LRWC and azimuth resampling, the squinted SAR data can be focused by exploiting a conventional broadside SAR imaging algorithm. Simulated and acquired data are studied to evaluate the method and assess its performance, and promising results are obtained.4. The proposed azimuth resampling based imaging algorithm is extended to process highly squinted spotlight FMCW-SAR data. For FMCW-SAR, due to the long time duration of the transmitted signal, the go-stop-go approximation which is widely used in the traditional SAR imaging algorithm derivation becomes invalid. To solve this problem, according to the dechirp-on-receive technique of FMCW-SAR and combining with the azimuth resampling processing, a suitable full aperture approach is presented for highly squinted spotlight FMCW-SAR. After applied LRWC and azimuth convolution, the Doppler frequency shift induced by the continuous antenna motion iscompensated in the range-Doppler domain. This step can avoid residual RCM in the range-compressed data. Then, azimuth resampling is performed to eliminate the azimuth-variation induced by LRWC and, thereby, to restore the “azimuth-shift invariance” property. Finally, range cell migration correction, secondary range compression, and azimuth match filtering are implemented to focus the data. Moreover, for wide-swath case, a subswath processing method is introduced to accommodate the range-variation of range cell migration. The effectiveness of the proposed approach has been verified by the simulation experimental results.5. A new SAR imaging system based on periodic non-uniform sampling is proposed for high-resolution wide-swath imaging. Due to the minimum antenna constraint, high-resolution and wide-swath becomes a contradiction in the spaceborne single-channel SAR system. By choosing a suitable PRF and applying orthogonal coding waveform, the range aliasing can be avoided. Using the non-uniform sampling which is carefully designed, makes the range blind distribute uniformly in the echo data. As such, the overlap of blind ranges is avoided. Besides, the method uses the periodic sampling to construct the equivalent multi-channel data, and uses the multi-channel unambiguous reconstruction method to reconstruct the signal’s spectrum. The proposed system can realize high-resolution wide-swath imaging for single channel SAR system and there is no blind area along range direction. Simulation results from point target have verified the feasibility and the effectiveness of the method.6. An improved shift-and-correlation(SAC) approach is proposed to meet the Doppler rate estimation in spotlight and sliding spotlight SAR focused with a full aperture algorithm. The range-Keystone transform is introduced into the original SAC algorithm to eliminate the coupling effect between the cross-correlation peak position and the target’s distance. Thus, the constraint of the focus depth induced by accumulating a number of detected cross-correlation functions along range can be avoided. Since there is no zero padding in the proposed approach, aliasing will occur in the Doppler rate estimation. For this problem, an ambiguity number estimation method based on minimum entropy is presented and the Doppler rate expression also is modified to get the right estimation value. The simulated and real spotlight SAR data processing results have validated the effectiveness of the algorithm.