Research on Active Detection Technology of Atmospheric Turbulence Intensity Based on Gaussian Beams and Gaussian Vortex Beams

Accurate detection of atmospheric turbulence intensity is essential for atmospheric optics and laser communication, yet efficient inversion methods remain insufficient. This work explores the feasibility of utilizing Gaussian beams and Gaussian Vortex beams for the inversion of atmospheric turbulence intensity (Cn2). We constructed an SLM-based turbulence simulation platform to generate phase screens with different turbulence intensities and acquire corresponding beam spot datasets. A ConvNeXt-based convolutional neural network is optimized, and its performance is compared with the traditional beam average width fitting method. Furthermore, a novel fusion inversion model is proposed by integrating beam average width parameters and beam spot feature information. Experimental results indicate that, for nine refined Cn2 turbulence grades, the optimized CNN achieves an inversion accuracy of 98%, while the beam average width method obtains 84.56% based on Gaussian Vortex beams. In contrast, the corresponding accuracies decreased to 95% and 73.84% when adopting conventional Gaussian beams. Compared with the standalone CNN model, the proposed fusion model achieves absolute accuracy improvements of 7.0% in indoor experiments and 2.0% in outdoor field tests. The results demonstrate that Gaussian Vortex beams combined with beam average width features exhibit superior performance in Cn2 inversion. The proposed fusion model presents reliable and promising inversion capability, which provides a feasible technical solution for the design and optimization of atmospheric turbulence monitoring systems.