Experimental Design for Improving Communication Performance of Vortex Beams Based on Neural Networks

A simulation experiment system based on a neural network for distortion compensation is designed to address the degradation in the transmission performance of vortex beams caused by atmospheric turbulence in free-space optical communication. Guided by the concept of education–research integration, a three-stage progressive structure of “basic characteristic analysis-turbulence impact assessment-neural network compensation” is adopted. First, using the Laguerre-Gaussian (LG) beam model, the intensity and phase distribution characteristics of different modal vortex beams are simulated and analyzed. Second, an atmospheric turbulence model is constructed based on the modified Von Karman spectrum and the random phase screen method, revealing the influence laws of turbulence intensity and propagation distance on beam quality. Finally, a U-net neural network is introduced to compensate for distorted beams by extracting turbulent phase information. Results show that the network can effectively identify phase distortions induced by turbulence of different intensities, with significant improvements in beam spot spreading and phase spiral structure deformation after compensation. This system is reproducible and extensible, providing a reference for experimental teaching and technological innovation in the field of optical communication.