The invention discloses an
underwater vortex
optical communication anti-turbulence method based on a photoelectric
hybrid deep neural network, and the method comprises the steps: carrying out the all-optical
information compression at a front end through the
diffraction deep neural network through the fusion of a
diffraction deep neural network and a
convolutional neural network, and transmitting the processed information to a rear end of the
convolutional neural network, a model task is completed through training, a predicted coefficient of a Zernike polynomial is output, prediction and compensation of ocean turbulence are achieved, and the
mode purity of vortex beams is improved. The front-end
diffraction deep neural network is utilized to effectively improve the problem that the rear-end
convolutional neural network is difficult to process massive information, the photoelectric
hybrid deep neural
network model is high in integration, high in operation speed and accurate in
light field signal processing, the power overhead for achieving turbulence resistance through the
machine learning technology is saved, operation is easy, implementation is easy, and the method is suitable for popularization and application. The method has a wide application prospect in the aspects of vortex beam
underwater communication and the like.