Sequence clean: A deconvolution algorithm useful for non-isolated radar target images with high sidelobes

High range-resolution radar imaging requires the use of large frequency spectra. However, large bandwidths suffer from various problems and present many obstacles. One is the desire to use the existing inventory of radar receivers that are wideband at rf but have relatively narrow if bandwidths. A second is the interference from jammers and other users of the spectrum. One of the ways to address these problems is to break up the large frequency band into smaller subbands, and obtain high resolution images after thinning the frequency spectrum. The first portion of the dissertation presents a detailed analysis of high resolution imaging with spectrally thinned wideband waveforms.; Spectral thinning results in high sidelobes in the range dimension. As a consequence, there is a reduction in the dynamic range and a deterioration in the image contrast. To improve the dynamic range and the image contrast, we propose a deconvolution technique called Sequence CLEAN (a variation of the CLEAN technique used in radio astronomy and Coherent CLEAN in microwave radar) for non-isolated radar target images suffering from high sidelobes. The latter half of this dissertation presents a detailed analysis of the various aspects of Sequence CLEAN, its susceptibility to noise (both system noise and the noise in the point spread function) and the computational costs. A demonstration with 2-D high resolution ISAR data is given which shows the performance of this algorithm in the range domain. Sequence CLEAN is applicable to the problem of high sidelobes both in range and azimuth.