Study On Airborne And Spaceborne SAR With Wide Swath

As an active microwave remote sensing technology for earth surface observation, synthetic aperture radar(SAR) has the characteristic of long distance illuminating under all-weather conditions day and night, which improves its capability of information acquisition. SAR plays important roles in military and civil applications, especially in the fields of battlefield reconnaissance, disaster forecast, resource exploration, terrestrial surveillance and deforming observation of ground surface and so on. High resolution and wide swath are always the motivation of SAR develpment. Some new teniques and imaging algorithms that can achieve wide swath for airborne and/or spaceborne SAR are studied in this dissertation.To achieve wide swath in single-channel spaceborn SAR systems, the antenna length in pitch dimension should not be very large as well as the pulse reputation frequency(PRF). Meanwhile, to achieve high resolution, the antenna length in azimuth dimension should be small, which increases the Doppler bandwidth. Thus, the azimuth signal may be aliased in the Doppler domain. Therefore, high resolution and wide swath contradict with each other in traditional single-channel spaceborne SAR systems.The modes of Scan SAR and terrain observation by progressive scans(TOPS) can achieve wide swath by sweeping the beams in pitch and azimuth direction, respectively. Different from Scan SAR mode, the phenomenon of grating lobe does not appear in TOPS mode by sweeping the beam in azimuth direction continuously. The zero-squinted TOPS mode has been successfully used in some spaceborne SAR systems, such as Terra SAR-X and sentinel-1. Combined with the squinted mode, TOPS SAR can illuminate a wide area forward the radar platform, which is very useful for air- and missile-borne platforms. Three main problems need to be solved in imaging processing for the squinted TOPS SAR. The first is original echo signal aliasing in the Doppler domain. The second is the severe coupling between the range and azimuth directions. The final is the aliasing in azimuth time domain after range cell migration correction(RCMC). These problems bring difficulties in imaging processing.With large transmitting power, increasing the beam operating range is another efficient way to achieve wide swath. For airborne strimap SAR systems with short antenna length in azimuth dimension, high resolution and wide swath can be obtained by increasing the beam operating range. However, the synthetic aperture length and range swath correspondingly increase, which will enlarge the sampling data drastically and thus bring more heavy storage and computation loads. On the other hand, long synthetic aperture length makes it hard to estimate the motion error of a full aperture. The above two problems make it difficult to design a real-time processing algorithm.The geosynchronous earth orbit(GEO) satellite operates at an orbit height of about 36,000 kilometres which is much higher than that of the low earth orbit(LEO) satellite. Thus, the swath of the GEO SAR is much wider than that of the LEO SAR. In addition, the revisit cycle of GEO SAR is only one day, which is very proper for In SAR. Therefore, GEO SAR is widely used in civil applications. However, high orbit leads to low relative speed as well as long synthetic aperture time. Thus, the Earth’s rotation cannot be ignored, which makes the relative motion between the antenna phase center and targets in the sence much more complicated than that of LEO and airborne SAR. Therefore, the related signal characteristics and imaging algorithms for GEO SAR should be studied.The dissertation focuses on the study of the problems mentioned above. Its main content are summarized as follows.1. A new range ambiguity suppression method for spaceborne SAR using azimuth multi-channel is proposed. Firstly, the characteristics of the azimuth phase coding are studied. After modulating the transmitted signal, the azimuth spectrum of the ambiguity signal is shifted, so that the spectrum of the useful signal and the ambiguity signal are not entirely superimposed. Then, the desired Doppler spectrum can be separated from the ambiguity spectrum using space domain filter which is constructed by muti-channel. Finally, conventional imaging algorithms can be applied to obtain image with wide swath.2. Three imaging algorithms for squinted TOPS SAR are studied. The echo of a sub-swath with full synthetic aperture is aliased in the Doppler domain. Using the azimuth preprocessing method, the two-dimensional spectrum without aliasing can be obtained. After the azimuth preprocessing, two full aperture imaging algorithms are studied. One is based on the azimuth frequency nonlinear chirp scaling(FNCS). Firstly, the two-dimensional spectrum without aliasing is obtained by the azimuth preprocessing after range walk correction. Then, the modified chirp scaling algorithm is used to complete RCMC. By using the FNCS method in the Range-Doppler domain, the variation of the chirp rates is corrected to achieve range compression. Subsequently, the azimuth signal is focused in the Doppler domain using the specan analysis(SPECAN) technique. Finally, geometric distortion correction is carried out. The other full aperture algorithm is based on the modified range migration algorithm(RMA) and the SPECAN technique. Firstly, the two-dimensional spectrum without aliasing is obtained by the modified azimuth preprocessing method. Then, the modified RMA is used to complete RCMC and range compression. Finally, the point targets are focused in the Doppler domain using the SPECAN and Dechirp techniques. Considering that the echo signal is unalised in sub-aperture, the third imaging algorithm for squinted TOPS SAR is studied. Firstly, the modified RMA is used to complete range compression and RCMC. Subsequently, the azimuth spectrum of the full aperture is obtained by sub-aperture recombination. Finally, the signal is focused in the Doppler domain by SPECAN and Dechirp techniques.3. A real-time imaging algorithm for zero-squinted airborne SAR with long operating range, wide swath and high resolution is studied. The motion error of the radar platform is largely compensated by the global position system(GPS) and interia measurement unit(IMU). After range compression, the data is divided into several blocks in range direction. The residual motion errors that are higher than second-order are estimated using the data of medium block, and then the motion errors of every blocks are compensated. Thus, the focused image with relatively low resolution can be obtained. The linear motion error can be estimated by the overlapped data between the adjacent sub-apertures and then the phase compensation funtion can be constructed. The imaging sence are divided into many grid cells. And the projection of every grid cells can be obtained by two-demensional interpolation in sub-aperture image. After compensating the phase, the projection data of every grid cells is coherent. With the increase of subaperture number, the azimuth swath becomes wider and wider and the azimuth resolution of point targets becomes higher and higher. Finally, the azimuth resolution of the point targets can achieve that of the full aperture data. The focused SAR image with wide swath and high resolution can be obtained by sub-image recombination in the range direction.4. Some main characteritics and several imaging algorithms for GEO SAR are studied. Firstly, the realistic geometric model of the GEO satellite and the Earth are constructed,which is used for the study of the relative motion between the antenna phase center and the ground targets. Based on this, the slant range mode is equivalent to the monostatic mode, which can simplify the signal model of the echo. Subsequently, the trajectory of the nadir, beam coverage, synthetic aperture time, Doppler characteristics and resolution of GEO SAR are studied. Finally, the back projection algorithm(BPA) and the singular value decomposition algorithm(SVDA) are applied to the echo signal simulation and imaging processing.