Recognition of Opposite-in-sign Orbital Angular Momenta of Laser Beams in a Turbulent Atmosphere with Neural Networks

An approach to increasing the communication channel capacity by amplitude and phase modulation almost reaches its limits. An increase in the information capacity of data communication channels by using the orbital angular momentum (OAM) of laser beams for information encoding is currently an urgent problem. The use of this approach in atmospheric optical communication systems is limited by the distorting effect of atmospheric turbulence, which makes decoding difficult and reduces the data transfer rate. In addition, the intensity distributions of vortex beams with OAMs opposite in sign are identical in a homogeneous medium, which also limits the use of OAM sign for encoding information. This work analyzes fundamental possibility of neural networks for recognizing opposite in sign OAMs of vortex beams in a turbulent atmosphere only by intensity distributions. The study is based on numerical simulation of Laguerre-Gaussian beam propagation in a turbulent atmosphere and use of the derived intensity distributions for training and testing neural networks. It is been shown for the first time that neural networks enables recognizing opposite in sign OAMs of Laguerre-Gaussian beams propagation through a turbulent atmosphere by intensity distributions with an accuracy of more than 90%. The results can be useful for developers and researchers of atmospheric optical communication systems where OAMs of vortex beams are used.