Machine Learning Corrects 2.4km Free-Space Optical Link Wavefront Errors, Reducing Phase Error Variance by 2/3

Atmospheric turbulence presents a significant challenge to free-space optical communication, distorting light signals and limiting reliable data transmission. Nathan K. Long from the University of New South Wales, alongside Benjamin P. Dix-Matthews, Alex Frost, and colleagues at the University of Western Australia, now demonstrate that standard assumptions about signal and reference beam distortion in such links are inaccurate.

The team experimentally reveals measurable differences in wavefront errors between these beams over a 2. 4 kilometre atmospheric link, and develops machine learning algorithms to correct for these relative distortions. This innovative approach achieves substantial reductions in phase error, and crucially, suggests a potential order of magnitude increase in secure key rates for emerging continuous-variable quantum key distribution technologies, paving the way for more robust and secure optical communication systems.

Machine Learning Corrects Kilometre-Scale Optical Turbulence Researchers demonstrate machine learning techniques to correct wavefront errors in a 2. 4km free-space optical link, achieving improved communication fidelity. The study addresses the significant challenge of atmospheric turbulence, which distorts optical signals and limits the performance of free-space optical communication systems.

The team developed a system that learns to predict and compensate for these distortions in real time, enhancing signal quality and extending communication range. This approach utilises a wavefront sensor to measure the distortions introduced by the atmosphere, and then employs a machine learning algorithm to calculate the necessary corrections to apply to the transmitted signal. The results show substantial improvement in signal quality, paving the way for more reliable and efficient free-space optical communication systems, particularly for applications requiring high bandwidth and secure communication, such as satellite communication and inter-building data transfer. In coherent optical communication across turbulent atmospheric channels, reference beacons are commonly used alongside information-encoded signals. This team presents experimental evidence of relative wavefront errors between polarization-multiplexed reference beacons and signals after transmission through a 2. 4km atmospheric link. They developed machine learning-based wavefront correction algorithms to compensate for these observed errors, utilising phase retrieval techniques, resulting in up to a two-thirds reduction in the observed error.

Wavefront Errors Mitigated by Machine Learning This research presents experimental results demonstrating relative wavefront errors between optical reference beams and weaker signal beams in a free-space optical communication system. The core question addressed is whether these errors degrade signal quality and if they can be mitigated using machine learning-based wavefront correction. The research focuses on the implications for continuous-variable quantum key distribution systems.

The team established a 2. 4km free-space optical link and multiplexed signals using multiple orthogonal Hermite-Gaussian modes. They observed that the reference and signal beams experienced different wavefront distortions due to atmospheric turbulence, a phenomenon termed relative wavefront error. A machine learning model was trained to correct for these relative wavefront errors. Experiments confirmed the existence of significant relative wavefront errors between reference and signal beams. The machine learning-based wavefront correction algorithm successfully reduced the total wavefront variance, achieving a reduction of up to a factor of two. This reduction in wavefront error translated to a reduction in effective excess noise in the quantum key distribution system, potentially leading to a significant increase in secure key rates. Increasing the number of Hermite-Gaussian modes to increase capacity also increased transmissivity, which in turn increased effective excess noise, but the wavefront correction helped to offset this increase. The research suggests that implementing wavefront correction algorithms can significantly improve the performance of continuous-variable quantum key distribution systems over turbulent free-space optical channels. The machine learning-based correction offers a way to mitigate the detrimental effects of atmospheric turbulence on optical communication. The study highlights the trade-off between increasing channel capacity and increasing noise, demonstrating that wavefront correction can help to balance this trade-off.

The team achieved up to a factor of two reduction in total wavefront variance and up to a 17% reduction in effective excess noise, while maintaining high coherent efficiencies, approximately 0. 98, even with the corrections.

Wavefront Correction Boosts Quantum Communication Range This research demonstrates a significant finding regarding coherent optical communication through turbulent atmospheric channels.

Scientists have experimentally observed relative wavefront errors between reference beacons and information-carrying signals, a phenomenon previously assumed negligible. Through a 2. 4km atmospheric link, the team established that these relative wavefront errors exist and developed machine learning-based wavefront correction algorithms to mitigate their impact, achieving up to a two-thirds reduction in phase error variance. The successful implementation of these algorithms has important implications for continuous-variable quantum key distribution. Analysis indicates that employing similar wavefront correction techniques could potentially increase secure key rates by an order of magnitude, substantially improving the efficiency and security of long-distance quantum communication. While the source of these relative wavefront errors remains unidentified, the team emphasises that determining the origin is not essential for the effectiveness of their correction scheme. The authors acknowledge that the impact of wavefront corrections on coherent efficiencies may be more pronounced across longer links or in conditions of stronger turbulence. Future work will likely focus on exploring these effects and further refining the algorithms to optimise performance in diverse atmospheric conditions, ultimately contributing to more robust and secure free-space optical communication systems. 👉 More information 🗞 Relative Wavefront Error Correction Over a 2.4 km Free-Space Optical Link via Machine Learning 🧠 ArXiv: https://arxiv.org/abs/2512.04460 Tags: