Research Progress Of Orbital Angular Momentum Modes Recognition Base On Machine Learning

Orbital angular momentum(OAM),its inherently infinite dimensions can greatly increase the capacity of optical communication and information processing in both classical and quantum regimes.The resolution of OAM mode accurately recognized by the receiver is the key to increase communication capacity.Here,using a convolutional neural network(CNN)approach with an improved Res Net architecture,accurate recognition of light’s OAM with high resolution and broad bandwidth is demonstrated which is based on petal interference patterns.A type of hybrid beam carrying double OAM modes is utilized to provide more degrees of freedom for recognition of the OAM mode,e.g.the fractional topological charge number l and the angular ratio n.Our studies show that when l ranges from 1 to 10 and 2.00 to 2.09,and n varies from 0.01 to 0.99,even with an interval down to 0.01,the recognition accuracy rate of OAM is essentially 100%,a remarkable achievement toward generation and identification of both super high resolution and broad bandwidth of fractional OAM modes.These results will dramatically expand the applications for the next generation CNN-based OAM optical communication.