Radar Coincidence Imaging Technique Research

Coincidence imaging recently has been one of the most attention-getting subjects in optical imaging study, which leads to broad investigation and wide-ranging discussion. Motivated by the peculiarities of coincidence imaging, the paper proposes radar coincidence imaging technique, which achieves superresolution within beam via randomly modulating radar wavefront. Such an imaging method does not require target relative motion, which presents advantages in processing relatively stationary targets and non-cooperative ones, potentially provides a supplementary imaging technique for the conventional SAR/ISAR imaging methods.The study of radar coincidence imaging(RCI) focuses on the imaging process, signal model, basis principles, super-resolutin imaging algorithms, image-reconstruction under model mismatch etc.The basic theory of RCI is firstly investigated, which covers the system framework, transmitted waveform, and the principle of super-resolution within beam. Classical coincidence imaging is then analyzed so as to provide the key point to achieve RCI in the radar signal model: to produce radar signals presenting time-space independence in detecting area. Based on the comparison between RCI and classical coincidence imaging, the identity of the imaging principle is comfirmed for the two methods. Then the multi-transmitting configuration and the waveform of orthogonality and timeindependence are proposed to achieve the mulducation of random wavefront. Such an approach is proved effective to produce the detecting signals presenting time-space independence. The comparison is conducted between RCI and the conventional SAR/ ISAR and antenna-array imaging methods, which futher demonstrates RCI can resolve targets within radar beam without the requirement of target relative motion. Simulation results illustrate that RCI can obtain high-quality images either for relatively stationary targets or for the maneuvering ones via single return pulse.Based on the study of spatial ambiguity function, chapter 3 is devoted to the RCI nominal resolution, waveform design, antenna geometry design, and anti-jamming performance. The RCI spatial ambiguity function is presented and the explicite expression of the nominal resolution is provided, which demonstrates the relationship between radar parameters and the resolution. In addition, the resolution under the zerocenter-frequency assumption is provided, which can quantitatively measure the relationship between antenna geometry design and RCI resolution. Then the waveforms modulated via stochastic signals and chaotic signals are proposed, which can naturally satisfy the RCI requirement for orthogonity and time-independence at the same time. The robustness of RCI is investigated based on the conventional jamming approaches. Because of the time-space independent detecting singals, RCI presents high stability under jamming which is theoretically and experimentally demonstrated.In Chapter 4, the image-reconstruction method is studied based on the parameterized model. The correlation method, which is the basic image-recovery method of classical coincidence imaging, cannot achieve high resolution for RCI because the time-space independent characteristic of the detecting signal is limited by microwave systems. The parameterized image-recovery method is thus proposed and the solvable conditions are analyzed. In addition, the step-recovery algorithm is proposed which will greatly decrease the computation complexity. This chapter also focuses on producing time-space independent signals in three-dimensional detecting region so as to perform the three-dimensional RCI based on the parameterized model.In Chapter 5, we concern the imaging algorithms under model mismatch. The cause of RCI model mismatch has two types, i.e. target motion and initial grid mismatch. Based on the analysis of signal model, three facters of model-mismatch are provided: the grid position bias, the time-space independent characteristic of detecting signals, and the target scattering-coefficient vector. The limitations of the current algorithm are analyzed, based on which two approaches of image-reconstructions are provided to overcome the model-mismatch impact. In consideration of the motion-induced mismatch, the compressive sensing(CS) recovery algorithms are employed which utilizes the sparsity restriction to diminish the influence of model-mismatch error. Furthermore, two imaging schemes based on motion parameter estimation and the joint estimation of multi-dimensional parameter are provided so as to process high-speed targets. With respect to the initial gird mismatch, the current optimize-gird algorithms are no longer applicable for the image recovery of RCI. Therefore, the correlationparamterized method is proposed which uses the correlation method to estimate the grid-mismatch error and then iteratively modifies the results of the parameterized method. Such an imaging method can achieve high resolution with fine imagery quality under model mismatch. Moreover, the imaging algorithm based on matched filter(MF) is proposed, combining the conventional radar imaging technique and RCI methods. The MF-based RCI method not only considerably impoves the processing efficiency but also provides prior information for the image recovery under model mismatch.The summarization and the future work of this dissertation are discussed in charpter 7.