Optical Encoding Model Based on Orbital Angular Momentum Powered by Machine Learning

Area of Science:

• Optical physics
• Information theory
• Machine learning

Background:

• Orbital angular momentum (OAM) offers unique properties for optical data transmission.
• Laguerre-Gaussian (LG) beams are key to OAM-based encoding.
• Efficient and robust optical encoding is crucial for advanced communication.

Purpose of the Study:

• To design a robust optical encoding model for efficient data transmission.
• To leverage orbital angular momentum (OAM) properties of Laguerre-Gaussian (LG) beams.
• To integrate machine learning for decoding and verify model robustness.

Main Methods:

• Developed an optical encoding model using the intensity profile of superimposed OAM-carrying LG modes.
• Utilized LG beam indices (p, ℓ) for data encoding.
• Implemented a support vector machine (SVM) algorithm for data decoding.

Main Results:

• Achieved a bit error rate (BER) of 10^-9.
• Demonstrated robustness with a signal-to-noise ratio (SNR) of 10.2 dB in one SVM model.
• Successfully decoded data using machine learning-based detection.

Conclusions:

• The proposed optical encoding model is robust and efficient for data transmission.
• SVM-based decoding effectively verifies the model's performance.
• OAM properties of LG beams provide a viable foundation for high-performance optical communication.