Fourier domain-based iterative predistortion method for high-repetition-rate optical linear frequency sweep

Abstract

The high frame rates and dense point clouds required by frequency-modulated continuous wave (FMCW) LiDAR necessitate high-repetition-rate laser frequency sweeps. However, limited research has addressed the predistortion of frequency-swept lasers (FSL) operating at high repetition rates. Here, we propose a Fourier-domain-based iterative predistortion (IPD-FD) method to achieve precise linearization of the swept laser frequency. This approach optimizes a sequence of Fourier series to generate the desired voltage waveform. To mitigate the end effects during instantaneous frequency extraction, a moving average technique is employed, significantly enhancing the stability of iterative convergence. We evaluated the predistortion method across varying repetition rates and sweep bandwidths, demonstrating that the mean 1-r² remains at the order of 10⁻⁷. Compared to temporal correction methods focused on specific regions of interest (ROI), the proposed method offers superior waveform matching across both ROI and non-ROI regions, thereby improving the overall stability of predistortion. Finally, ranging experiments conducted over a distance of 174 m yielded a precision of 4.8 mm. These results highlight the effectiveness of the IPD-FD method in linearizing FSL at high repetition rates, paving the way for its application in FMCW LiDAR systems requiring dense point clouds and high-resolution performance.