Review of vortex beam orbital angular momentum mode detection methods

Orbital angular momentum (OAM) is a new physical dimension that has revolutionized optical communication. This paper presents the common vortex beams (VBs) and the state-of-the-art methods for detecting their OAM modes. We show how machine learning can overcome the limitations of traditional OAM pattern recognition methods, such as their low accuracy and inability to handle fractional or mixed orders. We use neural networks to extract the features of the samples and perform pattern recognition and classification. We also discuss the application of deep neural networks in OAM pattern recognition, considering the neural network architecture and the beam propagation disturbance. Furthermore, we introduce the OAM coding encryption based on neural networks. This paper demonstrates how technological advances can improve the quality and speed of free-space optical communication systems.