Study On Moving Target Imaging Technique Of Single-Antenna Airborne SAR

As an all-day and all-weather advanced imaging modern sensor of high resolution, SAR has been widely used in many civilian and military applications. In SAR system, moving target imaging is an important application. On the hardware side, because single-antenna SAR is low-cost and does not exist registration problem caused by multiple antenna. And duo to a less amount of data storage, the SAR is easy to achieve real-time processing. In addition, most of airborne SAR systems still acquire data in single-antenna model. Therefore, there has been continuing interest in the research for moving target imaging in single-antenna SAR.Compared with stationary targets, moving targets need to face a much serious coupling between range and azimuth due to the existence of moving parameters. Particularly for fast moving targets, the coupling is more serious. So moving targets bring us not only information, but challenges as well.This thesis studied high-resolution imaging and moving parameter estimation of moving targets in single-antenna airborne SAR. Main works and innovations are summarized as follows:1. A general model of moving targets considering the altitude of platform is established. First, a conventional model of moving targets is analyzed. Then, a more practical model of moving targets is build. Comparing with the new model, the conventional model is found lacking. In the new model, range velocity through the Doppler modulation can affect the range curvature. Further, when the Doppler centroid is used to estimate range velocity, the slope distance of moving targets located in a focused image needs to be taken into account.2. An extraction algorithm based on maximum amplitude is proposed. In existence of large range migration, the features of maximum amplitude in image are used to locate targets’ position. In this way, on the one hand, single target can be separated from multiple targets. It is advantageous to obtain a high resolution image for multiple targets by processing each target individually. On the other hand, in order to realize phase unwrapping, according to the corresponding echo data of the target’ position, the phase of the target can be directly extracted. It can improve the performance of the subsequent Doppler parameter estimation algorithm. The contradiction between range migration and phase estimation can be resolved by directly obtaining the phase of the echo signal before range migration correction. For noise and clutter in scenes, some processing algorithms can be proposed and the feasibility of the algorithm is analyzed. Simulations demonstrate multiple targets separation and phase extraction can be achieved by using the extraction algorithm based on maximum amplitude in complex environments.3. A moving parameter estimation algorithm based on maximum amplitude is proposed. The Doppler parameters are the accurate corresponding relation with the moving parameters by analyzing the model. So the Doppler parameters can be used to estimate the moving parameters. For the presence of Doppler ambiguity caused by fast moving targets, Doppler ambiguity estimation algorithms in different noise intensities are proposed. In weak noise environment, the slope is used to estimate the ambiguity number efficiently. In strong noise environment, the contrast is used to estimate the ambiguity number accurately. In order to improve computational efficiency in strong noise, a more efficient adaptive and iterative ambiguous number estimation algorithm is proposed based on the maximum amplitude extraction algorithm, which considers adequately extracted information and is more efficient. Depending on the relationship between moving parameters and Doppler parameters in the model, moving parameters can be estimated. Simulations demonstrate a high resolution image of moving targets is obtained by using the Doppler parameter estimation algorithm, and moving parameters is estimated precisely by using the moving parameter estimation algorithm in strong noise environment.Moreover, a novel two-dimension Doppler parameter estimation algorithm is proposed based on maximum amplitude extraction by combining radar signal processing and image processing in this thesis. In the existence of large range migration, the algorithm is still valid and compensates the lack of conversational time-frequency analysis algorithm.