Cnn-based phase matching for the oam mode selection in turbulence heterodyne coherent mitigation links

Abstract

A novel method is proposed to select vortex beams carrying a specific orbital angular momentum (OAM) mode in turbulence heterodyne coherent mitigation (THCM) link. It is worth mentioning that intelligent phase matching (IPM) of the OAM beams based on the convolutional neural network (CNN) is the remarkable feature. Namely, CNN is particularly trained as the OAM modes classifier by the light intensity distribution patterns of different modes. The classifier actually acts as a mode detector to distinguish OAM modes by the map between the light intensity distribution and OAM mode, and then output mode information (MI). Specially, the phase matching technology is demonstrated to realize selection of specific OAM mode, where exploiting MI to select a specific phase mask is a characteristic of IPM. Subsequently, the phase mask is attached to the Gaussian beam to obtain the OAM beam carrying a special mode. Numerical results show a high IPM accuracy of 99% under medium strength atmospheric turbulence (AT).