Study On Key Techniques Of MIMO-SAR

Multiple-input multiple-output synthetic aperture radar(MIMO-SAR) has been receiving a lot of attention and been regarded as a key technique because of its superiority in high-resolution and wide-swath(HRWS) observation, ground moving target indication(GMTI), interferometry synthetic aperture radar(In SAR), etc.In this dissertation, the two key techniques of MIMO-SAR: waveform design and signal reconstruction and imaging algorithm, have been studied for four main applications of MIMO-SAR: HRWS imaging, In SAR, three-dimensional SAR, and SAR-GMTI.The main work of the dissertation is summarized as follows:In Chapter 2, the beam-space MIMO-SAR using a new multidimensional encoded waveform based on the two-dimensional(2-D) intrapulse beam steering is presented to achieve HRWS imaging and high signal to noise ratio(SNR). The SNR is an important performance index for judging the quality of SAR images. However, the SNR will decrease as the swath becomes wider and the resolution becomes higher, with other conditions constant. In this dissertation, a new multidimensional waveform encoding technique based on the two-dimensional intrapulse beam steering is presented to not only sufficiently improve the SNR but also enable a flexible power assignation and the multiple range resolution ability, which fully utilizes the whole plane-array to form 2-D compact sub-beam and steer it in both range and azimuth directions. The two kinds of range ambiguities and the Doppler ambiguity brought by 2-D intrapulse beam steering are discussed in this dissertation, and a 2-D ambiguity resolving algorithm is presented to separate and reconstruct the sub-beam signals, so that the signal can be processed by conventional SAR imaging algorithms. Furthermore, the system performances are also discussed, and the comparison between the proposed beam-space MIMO-SAR and three typical HRWS imaging modes is made, which shows that the proposed beam-space MIMO-SAR has a close swath size but a much higher SNR when compared to the three typical HRWS imaging modes. The simulation results validate the feasibility of the proposed multidimensional waveform encoding technique, and the effectiveness of a 2-D ambiguity resolving algorithm presented in this chapter.In Chapter 3, a class of azimuth phase coding linear frequency modulation(APC-LFM) waveforms are presented for MIMO-SAR application in elevation,such as In SAR and three-dimensional SAR, etc. In MIMO-SAR, it is necessary to separate the radar echoes corresponding to each transmit waveform. However, most existing orthogonal waveforms cannot be reliably separated from the received radar echoes, thereby degrading the performance of MIMO-SAR. As a solution to this problem, the APC-LFM is proposed in this chapter. After the APC phase demodulation, because of the APC shift effect in the Doppler domain, the overlapped echoes corresponding to different proposed APC waveforms arrive from different azimuth angles and, thereby, are separable using azimuth DBF. The APC-LFM waveforms can be used not only for MIMO-SAR in elevation to improve the performance of multibaseline In SAR and three dimensions SAR imaging, but also for beam-space MIMO-SAR in elevation to transform range ambiguity into Doppler ambiguity so that the azimuth DBF can be used to resolve the range ambiguity, and then the range ambiguity resolving no more suffers performance degradation from the elevation fluctuation of the illuminated scene. The simulation results validate the feasibility of the APC-LFM signal and the effectiveness of the proposed echo separation method based on azimuth DBF.In Chapter 4, a new and efficient three-step algorithm to reconstruct the sliding spotlight mode concurrent multiple subbands MIMO-SAR signal is proposed, which solves three key problems: the channel sampling position error caused by curved orbit, the increased total Doppler bandwidth due to the beam steering, and the range history differences between subband signals. Firstly, to correct the multichannel sampling position error caused by the curved orbit, an error correction approach is presented, which closely resembles the motion error compensation method. Secondly, the azimuth reconstruction based on a new multiple subband azimuth preprocessing is proposed both to reconstruct the subband signals and to keep the range history differences convenient to eliminate. Finally, a novel adaptable and efficient frequency-domain bandwidth synthesis method improves the frequency-domain bandwidth synthesis method, where an accurate frequency shift is performed without upsampling the signal.In Chapter 5, a new practical approach for ground moving target detection and imaging is presented for the squint mode two-channel MIMO SAR-GMTI system based on the forward-looking configuration. The proposed approach uses the squint mode range Doppler algorithm to focus the clutter, and then the slowly moving targets are toughly focused too, and can be detected and located. However, there are problems for the fast-moving targets: the migration through range cell as well as Doppler ambiguity in azimuth, which defocus the fast-moving targets in the SAR images and then the moving target detection and the parameter estimation cannot be performed. Furthermore, because the PRF is always slightly larger than the Doppler bandwidth, the Doppler spectrum folding always happens to the fast-moving targets, and causes a split of energy of fast-moving targets into two parts. Considering the facts that Doppler spectrum folding demands complicated processing, fast-moving targets after clutter suppression are reconstructed and refocused in order to detect these targets, in the procedure of which the Doppler spectrum is adequately compressed before range walk correction which effectively avoids Doppler spectrum folding and makes the procedure simpler.