Circular SAR processing using an improved omega-k type algorithm

Abstract

An improved circular synthetic aperture radar (CSAR) imaging algorithm of omega-k (ω-k) type mainly for reconstructing an image on a cylindrical surface is proposed. In the typical CSAR ω-k algorithm, the rage trajectory is approximated by Taylor series expansion to the quadratic terms, which limits the valid synthetic aperture length and the angular reconstruction range severely. Based on the model of the CSAR echo signal, the proposed algorithm directly transforms the signal to the two-dimensional (2D) wavenumber domain, not using approximation processing to the range trajectory. Based on form of the signal spectrum in the wavenumber domain, the formula for the wavenumber domain interpolation of the ω-k algorithm is deduced, and the wavenumber spectrum of the reference point used for bulk compression is obtained from numerical method. The improved CSAR w-k imaging algorithm increases the valid synthetic aperture length and the angular area greatly and hence improves the angular resolution of the cylindrical imaging. Additionally, the proposed algorithm can be repeated on different cylindrical surfaces to achieve three dimensional (3D) image reconstruction. The 3D spatial resolution of the CSAR system is discussed, and the simulation results validate the correctness of the analysis and the feasibility of the algorithm.