model-based image formation in ground-based circular strip-map synthetic aperture radar

one of the problems in the ground-based circular synthetic aperture radar is the dependence of the azimuth resolution to the antenna rotation radius and operational limitation of increasing it. to enhance the image quality in azimuth, model-based image formation through three methods: matched filter, inverse filter and optimum filter is suggested and analyzed. for the purpose of comparison, three quality measures for the noise, distortion and total error levels in the output image are extracted and it is showed that each of the three suggested image formation methods is optimum in one of these measures. for increasing the speed of the algorithm, the coefficient matrix of the proposed model is diagonalized by the replaced rows fourier transform matrix and using this property, it is showed that despite of accuracy, the proposed algorithm is faster than the well known back-projection algorithm.