Wavefront reconstruction of vortex beam propagation in atmospheric turbulence based on deep learning

Abstract

Wavefront distortion is one of the main challenges hampering the practical application of a vortex beam which carries orbital angular momentum. In this work, we propose a restoration method for wavefront distortion of the beam based on deep learning. The convolutional neural network model, which is capable of automatically learning the mapping relationship between optical field distribution of two kinds of vortex beams before and after the wavefront distortion, is well designed. After a large number of sample training, the CNN model possesses a good generalization ability to restore the distorted wavefront of the vortex beam by evaluating from mode purity, beam width and beam wander respectively under the conditions of different atmosphere turbulence, transmission distance and orbital angular momentum. Then, the experimental link is constructed to verify the effect of the model in the actual situation. The results show that the model proposed represents good robustness, which means the restored beam performs well in all indicators in different environments. The results provide a theoretical basis for the research of atmospheric turbulence suppression techniques of vortex optical communication.