| Synthetic aperture radar (SAR) has the capability to work all day and all weather and to obtain high-resolution images of the earth surface covered by the obstructions from a long distance. In recent years, with the development of the spaceborne SAR technology, a new type of remote sensing system, locating the SAR platform on a geosynchronous earth orbit (GEO) spaceborne satellite, has been becoming the central issue in the field of SAR imaging. Comparing with the low earth orbit SAR, GEO SAR has superior advertanges, such as short revisit period,wide swath, and high anti-strike ability, with broad application prospects of disaster management, ocean monitoring, intelligence detection, geographic surveying, and so on.Due to the high orbit height and long synthetic aperture time, GEO SAR has different characteristics from LEO SAR, for example, the collapse of “stop-and-go"assumption and the approximate of linear synthetic trajectory. Thus, the imaging methods used in LEO SAR, cannot achieve high focusing quality in GEO SAR. This dessertation has analyzed the characteristics of GEO SAR system in detail, and has done some further researches on GEO SAR zero-Doppler centriod control, echo signal modeling and high resolution imaging techonolgy. The main achievements are as follows1. Combining with the motion characteristics of GEO SAR, this dissertation has proposed a synthetic aperture time calculation and a resolution analysis method for GEO SAR's curved trajectory model. Firstly, the satellite orbit model is established based on the special six-coordinate systems, and the special movement characteristics are analyzed in detail. Based on the characteristic analysis, the relationship between the synthetic aperture time and satellite parameters has been studied, and the calculating method of GEO SAR spatial resolutions has been put forward based on the vector gradient method. The derived expression of the azimuth resolution is only related to the relative position between the satellite and target at the synthetic aperture initial and end moment, which is applicable to the curved trajecroty in GEO SAR. The proposed resolution calculation method for GEO SAR is the basis of following imaging algorithms.2. As for the problem of the time-varying Doppler centroid frequency caused by the Earth's rotation and elliptical orbit effects, this dissertation has put forward a 2-D zero-Doppler centroid control method for GEO SAR with elliptical orbit. At first, the Doppler parameters of GEO SAR has been analyzed in detail, presenting the mathematical expressions of the Doppler centroid frequency and the Doppler frequency rate, and the effects of the satellite's attitude errors on them have been discussed. Then, based on the vector opertaion, a new approach for zero-Doppler centroid control in GEO SAR, performing the arritude steering by a combination of pitch and roll rotation of GEO satellite platform, is put forward. This method can keep the Doppler frequency of echo signal closed to 0Hz. At last, a 2-D phase scan method is introduced to carry out the accurate compensation of Doppler centroid,which can avoid the difficulty of rotating the GEO SAR platform, improving the flexibility and reliability of zero-Doppler centroid control. Simulation results show that using the zero-Doppler cerntroid control method is an effective way to reduce the sensitivity of Doppler parameter errors and improve the image quality in GEO SAR.3. Study on the GEO SAR small-scene fast imaging algorithm. Firstly, based on the satellite orbital parameters, the slant range model for small-scene GEO SAR is established, and the applicability of the slant range model in GEO SAR is analyzed.Then, considering the “stop-and-go" assumption errors, this dissertation derives an accurate echo signal model, and proposes two algorithms for GEO SAR small-scene imaging. One is modified SPECEN imaging algorithm. This method has high imging efficiency and is suitable for real-time imaging processing. The other is improved NLCS algorithm. It can compensate the RCM that changes with range.This method is suitable for wide-swath scene imaging in GEO SAR. These two imaging methods can be considered as the extensions of SPECAN algorithm and NLCS algorithm from LEO/MEO cases to GEO cases. Finally, the point targert simulation results verify the effectiveness of the two imaging algorithms.4. Study on the GEO SAR large-scene and high-resolution imaging algorithm.Firstly, the high accurate slant range model for the curved trajectory in GEO SAR is established, and the applicable accumulating time for the slant range model is analyzed. Then, aiming at the problem that the target's RCM varing along the range and azimuth dicretions in large-scene GEO SAR imaging, this dissertation proposes two kinds of imaging algorithms. One imaging method is based on the extended Keystone transform. This method is a new data processing method to solve the problem of nonlinear RCM that the traditional Keystone transform that cannot deal with. It can be considered as the extension of the traditional Keystone transform.The other imaging method is based on the generalized Omega-k algorithm. This method contains a new Stolt processing method to solve the problem of the variant RCM along the azimuth direction that the Stolt processing method in the traditional Omega-k algorithm cannot deal with. The former has higher computational efficiency, and the latter has a wider range of application, which is suitable to achieve GEO SAR high-resolution imaging for larger sence. Finally, the simulation results verify the effectiveness of the proposed methods. |
PDF Full Text Request