Transmission Characteristics Analysis and Optimization of Coherent Structured Light Fields in Turbulent Atmosphere Based on Deep Learning

In order to address the attenuation, distortion and distortion problems of light field transmission in a turbulent atmosphere, a coherent structured light field analysis method based on deep learning is introduced, aiming to improve the anti-disturbance ability of the light field in atmospheric turbulence and optimize its transmission characteristics. By establishing an optimization framework combining a turbulent atmosphere model and DNN (Deep Neural Network), the transmission performance of coherent structured light fields under different turbulence intensities is studied. Firstly, a numerical model of turbulent atmosphere is established, and the transmission process of coherent structured light field under turbulence of different intensities is obtained through Monte Carlo simulation. Secondly, DNN is used to correct the distortion of the light field. The network is trained to enable it to recognize and optimize the spatial structure and phase information of the light field to reduce the impact of turbulence on the light field. The network model is trained using a large-scale dataset to ensure that it can adaptively adjust the characteristics of the light field under different turbulence conditions. Finally, the modified light field is controlled by real-time feedback combined with the optimization algorithm to maximize the transmission efficiency and recovery accuracy, thereby achieving optimized transmission of the light field in a turbulent environment. Under strong turbulence conditions, the MSE (Mean Square Error) without feedback control is the highest, which is 0.20-0.22, and the SSIM (Structural Similarity) is the lowest, which is only 0.68-0.70, and the transmission efficiency is 0.63-0.65; after real-time feedback control, the MSE drops to 0.15-0.16, the SSIM increases to 0.76-0.78, and the transmission efficiency improves to 0.74-0.75. The coherent structured light field optimization method based on deep learning provides an effective solution for light field transmission in turbulent atmosphere. By introducing deep neural networks and optimization algorithms, the transmission characteristics of the light field are successfully improved, providing technical support for optical communications and laser transmission in practical applications.