Research on Reconstruction Method of Aperture Synthesis Radiometer under Non-uniform Sample
|School||Huazhong University of Science and Technology|
|Course||Communication and Information System|
|Keywords||ASR Inversion NUFFT Non-uniform sample Robust statistics|
Aperture synthesis radiometer (ASR), which offers the benefit of high spatialresolution and instant image circle at cost of high complexity of signal processing andhardware, is an alternative to real aperture radiometer. In order to reduce the number ofredundancy baseline, further reduce cost and complexity of the whole ASR system, thecircular thinned array configuration and the rotary scanning circular thinned arrayconfiguration have been proposed. However, these array configurations suffer theextremely complex brightness image reconstruction process. As the sampling datameasured by these array configurations are distributed in a non-uniform grid, theconventional cartesian FFT method could not be used directly. The great distancebetween the sub-arrays of distributed ASR, lead to a baseline missing and visibility datanon-uniform distribution of the case. This paper focuses on the reconstruction method ofASR under non-uniform sample.Firstly, the status of ASR and study of non-uniform fast fourier transform ispresented. The principle of ASR imaging is introduced in detail subsequently. After that,we analysis the non-uniform sampling data measured by these different arrayconfigurations. This paper presents an non-uniform fast fourier transform(NUFFT)algorithm for imaging reconstruction of synthetic aperture radiometer. This algorithmrelies on a combination of the fast Gaussian gridding algorithm and the use of FFT on anoversampled grid. The numerical simulation results indicate that the proposed algorithmis accurate. Finally, the feasibility analysis of distributed ASR reconstruction method andthe analysis of baseline based on robust statistics is presented.