Lightweight wavefront prediction for atmospheric turbulence in free-space CVQKD systems

Satellite-to-ground free-space optical links enable space-based quantum key distribution (QKD) Satellite-to-ground free-space optical links provide an important platform for space-based quantum key distribution (QKD) networks. Continuous-variable quantum key distribution (CVQKD) is compatible with coherent optical communication systems, making it a promising approach for satellite-to-ground QKD. However, atmospheric turbulence introduces wavefront distortion and intensity fluctuations that degrade optical coupling and limit CVQKD performance. Conventional adaptive optics (AO) compensation may involve substantial computational complexity and latency, motivating efficient wavefront prediction for rapidly varying atmospheric channels. In this work, we propose a lightweight Att-ConvGRU-UNet framework for predictive wavefront compensation. The framework combines a multi-scale UNet for spatial feature extraction, a ConvGRU module for temporal modeling, and a convolutional block attention module (CBAM) for adaptive channel- and spatial-feature refinement. Residual learning is employed to predict phase increments and improve training stability. In addition to conventional wavefront metrics, single-mode fiber (SMF) mixing efficiency and secret key rate (SKR) are used as downstream metrics to evaluate the impact of prediction performance on CVQKD. The proposed model achieves a mean squared error (MSE) of 0.0030 rad 2 and a structural similarity index (SSIM) of 0.9865 on a three-layer dynamic non-Kolmogorov phase-screen dataset, with 0.342 M trainable parameters. Predictor-side latency is further evaluated at different direct prediction horizons to characterize the trade-off between prediction accuracy and computational time. This work provides a lightweight and computationally efficient predictive solution for atmospheric turbulence compensation in satellite-to-ground CVQKD, offering a promising approach for next-generation space-based quantum communication networks.

Authors

Publication Details

Journal
EPJ Quantum Technology
Published
2026-11-13
DOI
https://doi.org/10.1140/epjqt/s40507-026-00599-4
Primary Topic
Optical Wireless Communication Technologies
Type
article
Field-Weighted Citation Impact
0.00

Funders

Controls
|||
ALL TIME
JAN
FEB
MAR
APR
MAY
JUN
JUL
AUG
SEP
OCT
NOV
article

Lightweight wavefront prediction for atmospheric turbulence in free-space CVQKD systems

bozhi wei, Fuji Kuang, Lanjian Chen, Tian Zheng et al.
5 links
EPJ Quantum Technology
Optical Wireless Communication Technologies
article

Lightweight wavefront prediction for atmospheric turbulence in free-space CVQKD systems

bozhi wei, Fuji Kuang, Lanjian Chen, Tian Zheng, Chen Dong, Ao Liu
article en

Abstract

Satellite-to-ground free-space optical links enable space-based quantum key distribution (QKD) Satellite-to-ground free-space optical links provide an important platform for space-based quantum key distribution (QKD) networks. Continuous-variable quantum key distribution (CVQKD) is compatible with coherent optical communication systems, making it a promising approach for satellite-to-ground QKD. However, atmospheric turbulence introduces wavefront distortion and intensity fluctuations that degrade optical coupling and limit CVQKD performance. Conventional adaptive optics (AO) compensation may involve substantial computational complexity and latency, motivating efficient wavefront prediction for rapidly varying atmospheric channels. In this work, we propose a lightweight Att-ConvGRU-UNet framework for predictive wavefront compensation. The framework combines a multi-scale UNet for spatial feature extraction, a ConvGRU module for temporal modeling, and a convolutional block attention module (CBAM) for adaptive channel- and spatial-feature refinement. Residual learning is employed to predict phase increments and improve training stability. In addition to conventional wavefront metrics, single-mode fiber (SMF) mixing efficiency and secret key rate (SKR) are used as downstream metrics to evaluate the impact of prediction performance on CVQKD. The proposed model achieves a mean squared error (MSE) of 0.0030 rad 2 and a structural similarity index (SSIM) of 0.9865 on a three-layer dynamic non-Kolmogorov phase-screen dataset, with 0.342 M trainable parameters. Predictor-side latency is further evaluated at different direct prediction horizons to characterize the trade-off between prediction accuracy and computational time. This work provides a lightweight and computationally efficient predictive solution for atmospheric turbulence compensation in satellite-to-ground CVQKD, offering a promising approach for next-generation space-based quantum communication networks.

EPJ Quantum Technology
National Natural Science Foundation of China
Optical Wireless Communication Technologies
AI Navigator

Ask Laika to Summarize, Analyze, and Connect papers live on the map.

Summarize Papers & Methodologies

Extract key findings, datasets, and comparative methods across publications.

Benchmark Rankings & Visual Analytics

Rank top research institutions, authors, funders, topics, and journals by Field-Weighted Citation Impact (FWCI) and paper volume with instant charts.

Connect Distant Disciplines

Bridge topological clusters on the map to find hidden collaborative intersections.

Laika access coming soon