【作者】 蔡斌；

【导师】 梁甸农；

【作者基本信息】 国防科学技术大学 ， 信息与通信工程， 2009， 博士

【摘要】 分布式星载合成孔径雷达(SAR)是将卫星编队和星载SAR技术有机结合的新体制航天雷达系统。它通过多颗卫星编队飞行、协同工作,可实现干涉测高(InSAR)功能,使得快速、大范围地获取高精度DEM数据成为可能;可实现地面运动目标指示(GMTI)功能,为地表大面积监视、侦察和目标跟踪、定位提供了可能。该类系统具有重要的军事和民用价值,是目前国内外研究的热点。在提高性能的同时,分布式星载SAR在系统理论和技术上也存在诸多难点,面临许多挑战,需要新的系统设计理论及信号处理技术予以支撑。本文瞄准分布式星载InSAR和SAR-GMTI信号处理这一前沿课题,对分布式星载InSAR单/多基线信号处理方法和分布式星载SAR-GMTI单/多基线信号处理方法展开研究。论文各章节的研究内容安排如下:第二章研究了分布式星载InSAR单基线信号处理方法。首先,针对在复杂地形区域,星间长基线会导致干涉相位条纹密集、随机噪声增强的问题,提出了一种自适应多分辨率干涉相位滤波方法。该方法能有效地适应干涉相位二维空变特性,兼顾了抑制相位噪声和保持条纹细节间良好的综合性能,为获取高质量的DEM打下了基础。其次,提出了一种单基线InSAR叠掩和阴影区域检测方法。从叠掩和阴影区域的几何模型与信号模型出发,分析了叠掩、阴影和正常InSAR区域干涉信号相关矩阵的特点,提出了基于信息论的叠掩和阴影区域检测方法,在InSAR干涉相位中将叠掩和阴影区域隔离出来,保证了后续处理步骤的有效性和可靠性。第三章研究了分布式星载InSAR的多基线信号处理方法。以多星编队InSAR系统单航过为背景,研究了多基线条件下的叠掩、阴影和正常InSAR干涉区域的检测方法,以及叠掩区域的叠加信号个数估计、干涉相位估计和解缠方法。首先,采用基于信息论的干涉相位区域分类方法,将InSAR干涉区域分为叠掩、阴影和正常InSAR干涉区域,并估计出叠掩区域所包含的来自不同地面区域的信号个数。其次,研究了基于Root-MUSIC和RELAX的叠掩区域干涉相位估计与解缠方法,对叠掩区域中来自不同地面区域的信号分别进行相位估计和解缠。第四章研究了分布式星载SAR-GMTI单基线信号处理方法。首先,以双星单相位中心SAR-GMTI系统为背景,基于杂波和动目标的信号模型及统计分布模型,考虑杂波和加性噪声、通道幅度/相位不一致性误差、频率同步误差等影响因素,分析了SAR-ATI和SAR-DPCA技术的运动目标检测性能。其次,针对SAR-ATI技术的恒虚警检测问题,以“杂波”和“杂波+动目标”的幅相二维联合分布为基础,证实了选取幅度作为一维检测子的可行性,研究了相位单边、幅度单边和幅-相二级恒虚警检测方法;提出了基于条件分布的幅度-相位二级恒虚警检测方法和幅度-相位二维联合的恒虚警检测方法,并以基于NP准则的似然比检测方法的性能为上限,对以上恒虚警检测方法性能进行了分析和对比。最后,针对分布式星载单基线SAR-GMTI系统存在的混合基线问题,在地面场景为平地情况下,提出了修正的SAR-DPCA和SAR-ATI方法;在地面场景为起伏山区情况下,提出了一种自适应SAR-DPCA杂波抑制方法,取得了令人满意的结果。第五章提出了稳健的多基线SAR-GMTI自适应杂波抑制、运动目标检测、测速和重定位方法。首先,分析了分布式星载SAR-GMTI多基线信号处理中,杂波图像间去相关和运动目标导向矢量失配这两种重要的误差源,采用了对角加载技术和基于地面道路先验信息的动目标导向矢量估计技术,有效地抑制了这两类误差。其次,提出了基于地面道路先验信息的二元假设和多元假设恒虚警检测方法。再次,提出了一种基于地面道路先验信息的动目标测速和重定位方法,提高了估计精度,降低了运算量。第六章总结了全文主要工作、创新点及有待深入研究的问题。更多还原

【Abstract】 Distributed spaceborne SAR system is a novel spaceborne radar system combining satellite formation flying technology and spaceborne SAR technology perfectly. Through formation flying and cooperation of several satellites, the system utilize Interferometric SAR to measure the large-scale terrain digital elevation model(DEM) with high precision. Ground Moving Target Indication (GMTI) of distributed spaceborne SAR system gives the opportunity for large scale surveillance, spying, and target tracking, position. Because of its special value for military and civil application, the distributed spaceborne SAR system is currently one of the research focuses all around the world.The spaceborne SAR system faces a lot of system theory and technique challenges as well as it improves the performance. Since the system is a multilevel and complex large-scale system concerning broad technology, the novel system design theory and signal processing technique need to be studied to resolve the difficulties in the key techniques. This thesis focuses on the signal processing technique of InSAR and SAR-GMTI of the distributed spaceborne SAR system. The systematic research is carried out in single-baseline and multi-baseline signal processing technique of InSAR and SAR-GMTI in the paper. The research in each chapter is arranged as following:In chapter 2, the single-baseline signal processing techniques of the distributed spaceborne InSAR system are studied. Firstly, an adaptive multiresolution filtering approach for interferometric phase is proposed. In the complicate topographic fields, the long baseline between the two satellites makes the interferometric fringe densely and increases the dispersion of the phase noise. The proposed method can track the phase 2-D changes more accurately and achieve a better attention to removal of the noise and preservation of the topographic details. Secondly, a layover and shadow detection method is proposed for single-baseline InSAR. From the geometrical model and signal model of layover and shadow, the thesis proposed a layover and shadow fields detection technique based on the information theoretic criteria. The layover and shadow fields are separated from the interferometric phase to ensure the validity and reliability of the following processing steps.In chapter 3, the multi-baseline signal processing techniques of the distributed spaceborne InSAR system are studied. Based on the distributed spaceborne multi-baseline InSAR system, we study the detection of layover fields, the estimation of the number of the signal component, the estimation of interferometric phase and phase unwrapping. Firstly, based on the information theoretic criteria method, layover and shadow are isolated from the ordinary interferometric phase fields, and the number of the signal component in the layover fields are estimated. Secondly, interferometric phase estimation and phase unwrapping are studied based on Root-MUSIC and RELAX. In the layover fields, the signal from different fields are separated. Phase unwrapping are carried out in the ordinary interferometric phase fields and the layover fields separately.In chapter 4, the single-baseline signal processing techniques of the distributed spaceborne SAR-GMTI system are studied. Firstly, based on the signal model and statistical model of the two clutter cancellers, the detection performance of SAR-ATI and SAR-DPCA are analyzed and compared, with consideration to the channel amplitude-phase unbalance error, time/frequency synchronization error and the influence of clutter and noise. Secondly, based on the joint probability density function (PDF) of interferogram’s phase and amplitude of the two hypotheses“clutter”and“clutter plus signal”, the one-step constant false alarm rate(CFAR) detector with interferometric phase or amplitude and the two-step CFAR detector with marginal PDF of interferometric phase or amplitude are analyzed for their capabilities and limitations. A new two-step CFAR detector with conditional PDF and a new two-dimensional CFAR detector are proposed. The likelihood ratio test based on the Neyman-Pearson(NP) criterion is exploited as an upper bound for the performance of the above CFAR detectors. Lastly, the clutter suppression technique are studied based on distributed spaceborne SAR-GMTI system, which includes hybrid along-track and cross-track baseline. When the ground scenes are flat earth, the modified SAR-DPCA and SAR-ATI technique are proposed. When the ground scenes are fluctuant mountainous area, an adaptive SAR-DPCA technique is proposed to settle the problem under some conditions.In chapter 5, we proposed a robust optimum adaptive clutter suppression, moving targets detection, velocity measurement and target relocation technique for distributed spaceborne multi-baseline SAR-GMTI systems. Firstly, Some realistic problems related to the implementation of the processor are investigated, which include system multiplicative phase noise and steering vector mismatch. The diagonally loading techniques is resorted to improve the estimation of clutter statistics in the presence of multiplicative noise. A prior information, like road network information, is integrated into the optimum adaptive processor to reduce moving target steering vector mismatch. Secondly, two hypotheses and multiple hypotheses CFAR testing technique are proposed, which are based on the prior information of road network. Lastly, based on a priori information, the target velocity measurement and relocation technique are proposed, which improve the estimation accuracy with low computation load.In chapter 6, we summarize the major work of the thesis, the innovation point and the problems needs to be improved.更多还原