ROSAR imaging algorithm based on frequency analysis

In the existing imaging algorithms for synthetic aperture radar with rotating antennas (ROSAR), the methods based on the Taylor series expansion of the slant range suffer from mismatch problems in azimuth. In addition, the time-domain convolution method cannot rapidly correct the curvature. In the proposed ROSAR imaging algorithm in frequency domain, the echo was transformed into the two-dimension frequency domain directly using the stationary phase principle. In this case, spectral characteristics of the echo signal were retained completely. Thus, the correspondent decoupling function and matching function in frequency domain were constructed to achieve the range curvature correction and scene reconstruction. Simulation results demonstrate that the proposed algorithm can perform the curvature correction rapidly. Furthermore, the proposed algorithm can overcome the mismatch problems in azimuth effectively.