Research On Topology Optimization Of Antenna Arrays In Aperture Synthesis Microwave Radiometers

During the past few decades, there has been growing interest in the use of microwave radiometers for remote sensing of the Earth, space exploration, navigation, survillance and medical imaging. In order to provide high spatial resolution without requiring the very large and massive scanning antenna of a real-aperture system, aperture synthesis radiometers (ASR) are proposed. ASR can synthesize a large aperture by sparsely arranging a number of small aperture antennas and cross-correlating of these antenna pairs.As a crucial technique for ASRs, antenna array design plays an important role in radiometric imaging. In this thesis, several aspects of antenna array design in ASR are addressed, including different design criterions, new placement algorithms as well as corresponding array configurations, novel design of millimeter wave array antenna and so on. The main contributions of the thesis are presented as follows:An ant-colony-optimization-based method exploiting the general structure of low-redundancy linear arrays (LRLAs) is proposed, which can ensure obtaining large LRLAs while greatly reducing the size of the search space, therefore greatly reducing computation time. Based on the method, several analytical patterns for LRLAs are further derived, which can yield various array configurations with very low redundancy in nearly zero computation time.A combinatorial method for optimal thinned rectangular arrays is proposed. The completeness of (u,v) coverage provided by the thinned rectangular array constructed by the method is proved and the theoretical upper bound of the array's redundancy is derived, which is superior to that of U or T array.Two combinatorial methods for optimal thinned circular arrays are proposed to achieve the least redundancy in baseline length (modulus of baseline vector). One is half-circle thinning based on the restricted difference basis, the other is full-circle thinning based on the cyclic difference set (CDS). Both methods can greatly reduce the number of antennas and channels and provide analytical solutions for the element positions of thinned circular arrays.A particle-swarm-optimization-based method is proposed to optimize the element locations of the circular array for the uniform (u,v) coverage. By designing an effective objective function for the optimization procedure, the computational complexity is greatly reduced and more uniform (u,v) coverage is achieved compared to the previous methods in the literature.The minimum degradation array (MDA) is suggested for the optimum radiometric sensitivity of ASR (â–³TASR). Aiming at minimizing DF which characterizes the effect of redundant baselines onâ–³TASR, two methods are proposed to search for MDA:one is a greedy constructive algorithm; the other is a simulated-annealing-based method. Numerical comparison of both methods'performance is presented. To further improve DF especially for a densely populated array, a concept of augmented maximum baseline (CAMB) is proposed. Also, the lower bound of DF is derived. Simulation and experiment results demonstrate that the proposed MDA design can ensure the optimum sensitivity of ASR.A novel antenna array for ASR prototype is presented, which is a sparse antenna array with an offset parabolic cylinder reflector at millimeter wave band. The characterization of the whole antenna array architecture and each part of it are detailed. All measured results indicate that the antenna array performs well with narrow main beamwidth, low peak sidelobe level and small VSWR, which are all desired for passive imaging radiometer.Research on minimum redundancy array (MRA) is also extended to other fields like beamforming and MIMO radars. Antenna array design in minimum redundancy MIMO radars is addressed and two methods are proposed to achieve the minimum redundancy virtual array:one is the combinatorial method based on difference bases and CDSs; the other is the simulated-annealing-based method incorperating difference bases and CDSs as a-priori knowledge. Numerical studies show the effectiveness of both methods. Also, based on the Fourier properties of arrays with optimal coarray properties, the coarray properties of MRAs with minimum sidelobe level are presented for the first time. Further, the upper bound of the peak sidelobe level (PSL) of an MRA is derived.