Coherence shaping for optical vortices: a coherence shift keying scheme enabled by deep learning for optical...

Area of Science:

• Optics and Photonics
• Optical Communication Systems
• Information Theory

Background:

• Structured light-based shift keying techniques are crucial for meeting growing communication demands.
• Higher-order optical modes face challenges like diffraction divergence and susceptibility to perturbations.
• Existing methods require robust solutions for stable and high-capacity optical communication.

Purpose of the Study:

• To introduce a novel coherence shaping method for petal-like optical vortex structures.
• To enable the generation of non-diffraction interference states between coherent and incoherent states.
• To develop a deep learning-enabled coherence shift keying (CSK) scheme for enhanced communication.

Main Methods:

• Experimental demonstration of coherence shaping for optical vortices.
• Generation of tunable interference states by controlling coherence.
• Implementation of a deep learning model for interference state recognition in CSK.
• Experimental validation of system performance and bandwidth.

Main Results:

• Successful generation of non-diffraction interference states using coherence shaping.
• Deep learning model achieved high recognition accuracy (>0.997) for interference states.
• Minimum achievable visibility-level bandwidth confirmed to be 0.02.
• Demonstrated robustness in complex environments.

Conclusions:

• The developed coherence shaping method offers a new platform for optical communication.
• Deep learning significantly enhances the reliability and accuracy of CSK systems.
• The proposed system supports high-capacity and encrypted communication using low-order structured light modes.