Study On New Methods For Airborne Multichannel SAR-GMTI Processing

Due to its all-weather, all day /night and long range, Synthetic Aperture Radar (SAR) can enhance radar's information acquisition capability greatly, thus exhibits great value in both civilian and military applications. Since traditional SAR radars only simply take"pictures"of a stationary scene at different times. It is hard to measure motion in the scenes because the ground moving targets are usually embedded in the imaged stationary scene, i.e., clutter. In order to estimate the target's true position and velocity parameters accurately, clutter contamination of the target's signal must be minimized or even cancelled. If only the conventional SAR formation algorithms are utilized without considering the target's Doppler information, the target's image would be seriously distorted. Since two or more antennas (channels) can provide more degrees of freedom, multichannel airborne along-track SAR has powerful ability to suppress clutter and distinguish targets moving on land or water from surface clutters over a large area. Thus it has been used extensively in air-to-ground surveillance and reconnaissance.In recent years, many countries in the world have been making great efforts to develop spaceborne / airborne multi-channel ground moving target indication (GMTI) systems, study new GMTI theories and explore high efficient detection and location algorithms. This dissertation is supported by Advanced Defense Research Programs of China (No. 102060302) and National Science Foundation of China (No.60502044 and No. 60725103). It analyzes the airborne SAR-GMTI research development of home and abroad and represents detailed research on parameter design, channel balancing, slowly (fast) ground moving target detection and location, based on conceptual analysis, theory derivation and real data validation.The main content of this dissertation is summarized as follows.1. The design of spaceborne / airborne multi-channel SAR-GMTI basic parameters is studied. The constraint relation among the parameters is deduced and illustrated by an example. The basic performance for GMTI is analyzed and an example of the design parameters for different GMTI systems is also illustrated when given the wavelength and velocity.2. This dissertation develops a new channel balancing algorithm for coherence improvement applied to the multichannel airborne along-track synthetic aperture radar (SAR), which can be used for GMTI by comparing two SAR images obtained from different channels. The turbulent aircraft motion, beam pattern difference and frequency response difference which may degrade the coherence are discussed in detail. And the corrections of amplitude and phase errors are applied as the two-dimensional (2-D) adaptive filters, the first in the range-Doppler domain and the second in the range-frequency azimuth-time domain. It is shown that the new channel balancing algorithm can improve the coherence of SAR complex image pairs greatly and can simplify the clutter suppression methods. The effectiveness of the new algorithm is demonstrated with real data acquired by a Chinese X-band multichannel airborne SAR- GMTI system.3. Usually, the range migration through resolution cells makes the fast moving target hardly be detected due to serious defocusing. Additionally, PRF ambiguity strongly impairs accurate position estimation as a result of Doppler spectrum folding. A method for ground fast-moving target detection,position and imaging based on the tri-channel airborne SAR-GMTI is presented when the above problems occur. Firstly, DPCA is applied to range compressed and azimuth-uncompressed data for clutter suppression. Then a combination of the Canny,Ratio operators and the Hough transform is applied for moving tracks extraction. After that, a method based on two-order Keystone transform is applied for target imaging, detection and motion parameter estimation after resolving PRF ambiguity. Finally, the slow and fast-moving targets are together labeled on the SAR image. The results of raw data show the efficiency of the proposed method in the dissertation.4. A new method based on joint pixel interferograms is proposed for three-channel GMTI. Firstly, a joint probability statistical model is established using the statistical properties of multilook SAR interferograms after joint pixel processing. Then the false and leak alarms are analyzed. Meanwhile, considering the velocity ambiguity in the range of minimum detectable velocity (MDV), an energy center method is presented to resolve ambiguity. Lastly, the interferometric magnitude and unambiguous phase are jointed together to detect moving targets. The validity of new detection method is demonstrated by the results from real data.5. Ground moving target indication is one of the most important applications in a bistatic synthetic aperture radar (SAR) system as well as in a monostatic system. In the dissertation, an algorithm for detection and location of moving targets is presented with an azimuth-invariant bistatic multi-channel SAR, which consists of one transmitter (channel) and multi-receivers (multichannel). The algorithm is based on discussion of the particularities of the bistatic SAR configuration, including coherence improvement and clutter characteristics. And the corresponding compensating methods including two- dimensional (2-D) range-azimuth prefiltering and bistatic differential range correction (DRC) are proposed to solve these particularities. It is shown that, using the compensating methods, the stationary clutter can be suppressed and the moving parameters of ground targets can be accurately estimated. Finally, simulation results demonstrate the effectiveness of the proposed algorithm.