Accurate range profile alignment method based on minimum entropy for inverse synthetic aperture radar image formation

Accurate range profile alignment is an essential step for achieving the subsequent azimuth coherent processing for inverse synthetic aperture radar image formation. In this study, an iterative method based on minimum entropy is first proposed for accurate range profile alignment, which constructs a series of local quadratic curves to gradually approach the extremum of the entropy of range profiles. The accuracy of range profile alignment depends on whether the iteration could be convergent to the optimal extremum. Unfortunately, the entropy of misaligned range profiles usually has numerous local extrema. The low-pass profile filtering is able to smooth the entropy surface and eliminate the local extrema. Nevertheless, the detailed features of range profiles are also lost due to the filtering so that the alignment accuracy could be reduced. Finally, a circulation cascade processing by appropriately combining the proposed iterative method and the low-pass filtering is presented to make the range profile alignment both convergent and accurate. Simulations and real data are used to validate the performance of the proposed method on iteration convergence and alignment accuracy.