Study On Spaceborne Interferometric SAR Signal Processing

Spaceborne Synthetic Aperture Radar Interferometry (INSAR) successfullyintegrats the inteferometry theory and the Synthetic Aperture Radar (SAR) imagingprinciple. The spaceborne INSAR takes the advantage of the SAR system parameters,the attitude parameters, the geometry of orbits, and so on to survey any point on thesurface of the earth. Spaceborne INSAR is one of the hot issues in the SAR researchfield in recent years. Now, the efficient and accurate Spaceborne INSAR signalprocessing and simulation are the major problems in this field. In this thesis, somesignal processing methods are studied to improve the processing efficiency anddecrease the processing errors. On the other hand, the spaceborne INSAR simulationis also studied in order to understand the essence of the INSAR technique.First of all, the spaceborne SAR imaging rationale, imaging algorithms, and thespaceborne imaging geometry model are established and analyzed. Secondly, thespaceborne INSAR signal processing including the registration, phase unwrappingand the DEM inversion are deeply studied and invalided using the ERS-1/2 raw dataand the simulated spaceborne INSAR data. Some improved methods are suggested toincrease the INSAR process efficiency. Thirdly, INSAR by small satellites formationflying is studied. The baseline design method and the elevation precision allocationmethod are proposed. The simulation results validate the effectiveness of thesuggested methods.The major work and the innovations of the thesis are summarized as follows:(1) The spaceborne SAR imaging algorithms are studied systematically, theRange-Doppler, Chirp-Scaling and theω-k algorithms are analyzed in detail.The advantages and the disadvantages of them are compared according to thepoint imaging of ERS-1/2 parameters. The images of Rome and Greenland areobtained using the raw data of ERS-1/2, with the resolution of 25m×25m. (2) According to the spaceborne INSAR geometry, an INSAR mathematicalgeometry model, named PAM model (Projection Approximate Method) isproposed to describe the INSAR geometry and the echo characteristics. Thephase information and the correlation of the system under the PAM model arecomprehensively analyzed.(3) The INSAR registration algorithms are reviewed. These algorithms cannot solvethe problem of the efficiency of large image registration is considered. Therefore,the Hierarchical Registration method is proposed to solve this problem. The HRmethod adopts the hierarchical thoughts, processing the image parallel, whichimproves the pixel registration and the process speed. Secondly, the phaseunwrapping is the most important step in the INSAR process. The present phaseunwrapping algorithms, the most important step in INSAR processing, areintroduced and discussed. According to the network algorithm, an improvedmethod, named Parellel Network Flow phase unwrapping method is put forwardto increase the phase unwrapping efficiency. The method is adapted to the largeimage. The practical Rome and Greenland data validated the HR and the PNFmethods. Thirdly, according to the unwrapped phase and the flat earth phase, theabsolute DEM is obtained.(4) The spaceborne INSAR scene simulation is studied. The Kirchhoff approximateapproach and the real scene model put forward by Franceschetti are adopted torealize the raw signal simulation of natural ground in a large area with thegeometry between the facet and the SAR. The spaceborne distributed raw datasimulation method and the satellites orbit simulation method are also realized.According to the ERS-1/2 orbit parameters and the DEM, the three dimensionspaceborne SAR distributed objects are successfully simulated. By theinterferomtric process, the interferometric images and the interferometric stripesare obtained, which validate the simulation method of spaceborne INSAR naturalscene distributed objects. The experiments prove that the method can simulatethe INSAR baseline decorrelation effect.(5) The small satellite formation flying is the new approach to implement spaceborne INSAR. The small satellite formation flying system schemes is introduced anddiscussed. The feasibility of the schemes is analyzed in detail. A baseline designmethod of small satellite formation flying is proposed. The two SAR satellitesformation flying system working in C band is analyzed by this baseline designmethod. And the elevation precision allocation method is also given.