Polarized deep diffractive neural network for classification, generation, multiplexing and de-multiplexing of orbital angular momentum modes
This addresses a critical bottleneck in optical communication for multiplexing and de-multiplexing polarized OAM modes, though it is incremental as it builds on existing diffractive neural networks.
The paper tackled the problem of handling orbital angular momentum (OAM) beams with varying polarization in optical communication by proposing a polarized optical deep diffractive neural network, achieving successful classification of 14 orthogonally polarized vortex beams and de-multiplexing hybrid OAM beams into Gauss beams at multiple spatial positions.
The multiplexing and de-multiplexing of orbital angular momentum (OAM) beams are critical issues in optical communication. Optical diffractive neural networks have been introduced to perform classification, generation, multiplexing and de-multiplexing of OAM beams. However, conventional diffractive neural networks cannot handle OAM modes with a varying spatial distribution of polarization directions. Herein, we propose a polarized optical deep diffractive neural network that is designed based on the concept of rectangular micro-structure meta-material. Our proposed polarized optical diffractive neural network is trained to classify, generate, multiplex and de-multiplex polarized OAM beams.The simulation results show that our network framework can successfully classify 14 kinds of orthogonally polarized vortex beams and de-multiplex the hybrid OAM beams into Gauss beams at two, three and four spatial positions respectively. 6 polarized OAM beams with identical total intensity and 8 cylinder vector beams with different topology charges also have been classified effectively. Additionally, results reveal that the network can generate hybrid OAM beams with high quality and multiplex two polarized linear beams into 8 kinds of cylinder vector beams.