Research On Key Techniques Of Ground Moving Target Indication For Synthetic Aperture Radar

Synthetic aperture radar (SAR) has the capability to monitor large areas of interest and therefore has been widely used in mapping the geological structures and features of a stationary scene. These advantages make a convincing case for their construction. Ground moving target indication (GMTI) is a hot topic in the area of SAR and plays an important role in both military and civilian field. The use of the SAR system in civilian and military areas has attracted the attention of researchers all over the world for application such as ground moving targets detection, imaging and positioning.For SAR-GMTI system, due to the high platform velocity, the Doppler bandwidth of the clutter will be expanded. The ground slowly moving targets are usually embedded in the strong clutter background. In order to detect the moving targets, the clutter should be well cancelled. Therefore, clutter suppression becomes a key technique in the GMTI community. In SAR images, the moving targets are displaced and blurred when the traditional SAR imaging algorithm is applied to a scene with moving targets. In order to accurately obtain the target's true position and velocities, the motion parameter estimation is also very important.In this dissertation, ground moving target detection and motion parameter estimation algorithms are addressed. Among the contributions are:1. The design of SAR-GMTI basic parameters is illustrated. The constraint relation among the parameters is analyzed. The relationship between the parameters and the performance of ground moving target detection is discussed. These solid results can provide some considerations for SAR-GMTI system design.2. A multi-channel ground moving target detection method is proposed based on the covariance matrix eigen-decomposition. The variation of the small eigenvalues summation of the covariance matrix is adopted to detect moving targets. The target radial velocity is estimated by the following two steps. Firstly, use the interferometric phase of the two SAR images to get the coarse velocity estimation, and then obtain the more precise value by searching the target spatial steering vector. This method can overcome the influence of the interferometric phase caused by clutter and noise.3. A clutter suppression method is proposed for heterogeneous environment. Adaptive filtering is an effective method for clutter suppression, whereas the performance degrades severely in heterogeneous clutter environment. To mitigate this problem, a new clutter suppression method is presented. This approach uses the subspace tracking technique to update the clutter subspace of different clutter patches, and then can relieve the heterogeneous effects. Simulation results illustrate that the method performs well when the independent and identically distributed (i.i.d.) samples is lacking.4. Two target parameter estimation algorithms are proposed. The first algorithm is based on contrast optimization criterion. The estimation of the velocities is obtained by searching the peak magnitude of the contrast function. The moving target parameter estimation method is analyzed in detail and it can enhance the estimation accuracy obviously. The algorithm is successfully applied to real SAR data. The second algorithm is proposed to overcome the shortcomings of the traditional moving target parameter estimation methods that assume the target moves with a constant velocity. The estimation of the moving parameters can be obtained by fitting the instant frequency and solving the phase history equations. Furthermore, by reconstructing the azimuth reference function, the focused images of targets can be obtained.5. Aiming at the distributed SAR systems with InSAR formation, the influence of the cross-track baseline on the performance of target detection are analyzed. The ideal formation for moving target detection is along-track, but one important mission of distributed satellite is to retrieve the digital elevation model (DEM). Therefore, the cross-track baseline component is inevitable. The cross-track baseline of the distributed satellites is sensitive to the terrain fluctuation which may badly affect the performance of clutter suppression. We analyzed the clutter suppression capability with different cross-track baselines and different terrain circumstance. Simulation results provide some basis to the satellite formation design.