Free-Space Optical Channel Turbulence Prediction: A Machine Learning Approach

2024-05-27

arXiv.org

AI Summary

Atmospheric turbulence degrades channel performance in free-space optical communication, yet conventional turbulence monitoring requires dedicated hardware. Method: This paper proposes a hardware-free turbulence-level prediction method that operates solely on the raw received optical bitstream—eliminating the need for specialized turbulence sensors—and systematically investigates the impact of temporal scale on classification robustness using time-series bitstream features. Contribution/Results: Employing multiple machine learning classifiers with hyperparameter optimization, the method achieves >98% classification accuracy across six controlled turbulence levels in a laboratory setting. To our knowledge, this is the first work to achieve high-accuracy turbulence-level identification exclusively from the received optical signal, establishing a novel paradigm for lightweight, low-cost adaptive optical communication systems.

Technology Category

• Machine Learning: Hardware-aware ML
• Search and Optimization: Learning to Search
• Computer Vision: Learning & Optimization for CV

Application Category

• Systems and Infrastructure for Web, Mobile and WoT: Applied ML and AI for Web-based mobile applications
• Security and Privacy: Large-scale security measurements
• User Modeling, Personalization and Recommendation: On-Device user modeling, personalization, and recommendation

Abstract

Channel turbulence presents a formidable obstacle for free-space optical (FSO) communication. Anticipation of turbulence levels is highly important for mitigating disruptions. We study the application of machine learning (ML) to FSO data streams to rapidly predict channel turbulence levels with no additional sensing hardware. An optical bit stream was transmitted through a controlled channel in the lab under six distinct turbulence levels, and the efficacy of using ML to classify turbulence levels was examined. ML-based turbulence level classification was found to be>98% accurate with multiple ML training parameters, but highly dependent upon the timescale of changes between turbulence levels.

Problem

• Predicting turbulence in FSO without extra hardware
• Classifying turbulence levels using machine learning
• Achieving high accuracy in turbulence level prediction

Innovation

• Machine learning predicts turbulence without hardware
• ML classifies turbulence levels with 98% accuracy
• Turbulence classification depends on timescale changes