Phase Encoding of Fractional Orbital Angular Momentum for high-capacity Optical Communication

Phase encoding is usually ignored in previous orbital angular momentum (OAM) studies, as it does not affect the OAM's intensity distribution. However, the intensity distribution of a fractional orbital angular momentum mode (FOAM) exhibits a radial gap, which rotates to different direction for different phase encoding. In this paper, we propose a high-capacity optical communication scheme based on the phase encoding of FOAM. In the scheme, the binary information is encoded on both the topological charges (TCs) and phases of a FOAM submode. All encoded sub-channels are transmitted through the communication channel. At the receiver, a deep learning recognition method based on convolutional neural networks (CNN) is employed for interference-free decoding. The experiments demonstrate that one FOAM sub-mode can transmit 4 bits by both the phases and the TCs. A high-capacity system with sixteen sub-channels multiplexing and demultiplexing is demonstrated (total: 64 bits), with the bit error rate (BER) being lower than 2 ×10
^-4
.