A new reconfigurable methodology to implement the shift-and-correlate (SAC) algorithm for real-time SAR autofocusing

Doppler Frequency Rate (FDR) is an essential parameter in the procedure of Synthetic Aperture Radar (SAR) processing. In this paper, we investigate an achievable autofocus algorithm, shift-and-correlate (SAC), and design the implementation of FDR estimation. The proposed method can dramatically improve the accuracy of imaging and reduce the time delay of the SAR imaging process. Depth-of-focus theory and linear fitting theory are discussed in this paper for decreasing the computation-consumption and the preserving precision. We verify the proposed method in a 16384*16384 SAR imaging process at 100 MHz system clock on a dedicated designed Field Programmable Gate Array (FPGA) tested. To testify the feasibility of the proposed method, we compared our run-out results with existing implementations using traditional DSP and MATLAB methods in terms of real-time performance and imaging quality.