Novel compressive sensing-based Dechirp-Keystone algorithm for synthetic aperture radar imaging of moving target

摘要: The authors propose a novel compressive sensing (CS)-based Dechirp-Keystone algorithm (DKA) for synthetic aperture radar (SAR) moving target imaging, which is called the CS-DKA. The DKA can focus on moving targets in range-Doppler domain efficiently through only keystone transform (KT), complex multiplication and Fourier transform (FT)/inverse Fourier transform (IFT) operations. It has been shown that the non-interpolation implementation of KT can be expressed by an orthonormal basis, and it is known that the complex multiplication and FT/IFT are linear and invertible; therefore, the Dechirp-Keystone operator (DKO) is also linear and invertible. In the proposed algorithm, the authors take the inverse of DKO (IDKO) rather than the exact SAR echo model to construct the representation basis in the CS frame owing to its high implementation efficiency. After that, a random transmitting/receiving scheme is considered, to implement the down-sampling operation, and then reconstruct the moving target image by solving a regularisation problem. Both simulated and real SAR data are processed to show that the CS-DKA with down-sampled data can focus the target as well as the conventional DKA does with full data, and at the same time can achieve much lower sidelobes.