An Efficient Range Scaling Algorithm for MIMO-SAR 3-D Imaging

In recent years, multiple-input-multiple-output synthetic aperture radar (MIMO-SAR) has been an effective scheme for millimeter-wave (MMW) 3-D imaging. Due to the extremely large amount of data in MIMO-SAR, fast imaging algorithms for it have been extensively studied. However, current imaging algorithms for such scheme are unsatisfactory, either being slow with high accuracy or introducing a large number of approximations that compromise imaging accuracy. In this article, we propose an efficient range scaling algorithm (RSA) to fetch up these shortages. We derive the theoretical formulation of RSA applied to MIMO-SAR, which can realize 3-D reconstruction without interpolation and accumulation operations, only by matrix multiplication and fast Fourier transform (FFT). Meanwhile, we analyze the computational complexity and computational error of the algorithm, and compared with the current advanced imaging algorithms (RMA, FSA, etc.), the proposed algorithm has higher imaging speed and is easier to realize real-time imaging. We also propose a new matrix rearrangement method which makes the algorithm more concise and faster. Based on the MIMO-SAR prototype system, experimental validation is carried out in the 75-110 GHz band. Experimental results show that the proposed algorithm has excellent efficiency and accuracy.