Quantum Wavefront Correction via Machine Learning for Satellite-to-Earth CV-QKD

State-of-the-art free-space continuous-variable quantum key distribution (CV-QKD) protocols use phase reference pulses to modulate the wavefront of a real local oscillator at the receiver, thereby compensating for wavefront distortions caused by atmospheric turbulence. It is normally assumed that the wavefront distortion in these phase reference pulses is identical to the wavefront distortion in the quantum signals, which are multiplexed during transmission. However, in many real-world deployments, there can exist a relative wavefront error (WFE) between the reference pulses and quantum signals, which, among other deleterious effects, can severely limit secure key transfer in satellite-to-Earth CV-QKD. In this work, we introduce novel machine learning-based wavefront correction algorithms, which utilize multi-plane light conversion for decomposition of the reference pulses and quantum signals into the Hermite-Gaussian (HG) basis, then estimate the difference in HG mode phase measurements, effectively eliminating this problem. Through detailed simulations of the Earth-satellite channel, we demonstrate that our new algorithm can rapidly identify and compensate for any relative WFEs that may exist, whilst causing no harm when WFEs are similar across both the reference pulses and quantum signals. We quantify the gains available in our algorithm in terms of the CV-QKD secure key rate. We show channels where positive secure key rates are obtained using our algorithms, while information loss without wavefront correction would result in null key rates.

💡 Research Summary

The paper addresses a critical limitation in satellite‑to‑Earth continuous‑variable quantum key distribution (CV‑QKD): the assumption that reference pulses and quantum signals experience identical wavefront distortions after propagating through the turbulent atmosphere. In practice, a relative wavefront error (WFE) can arise between the two, degrading the coherence between the real local oscillator (RLO) and the quantum signal and potentially nullifying the secure key rate.
To overcome this, the authors propose a machine‑learning‑driven wavefront correction scheme that leverages multi‑plane light conversion (MPLC). An MPLC decomposes the incoming optical field into a set of Hermite‑Gaussian (HG) modes, yielding per‑mode phase measurements for the reference pulses (Δϕ_mn,R) and, in principle, for the quantum signals (Δϕ_mn,S). Since measuring Δϕ_mn,S directly would require additional hardware, the authors train a transformer‑based neural network to map Δϕ_mn,R onto an estimate of Δϕ_mn,S.
Training data are generated via Monte‑Carlo simulations of a phase‑screen atmospheric model that incorporates altitude‑dependent turbulence, Fried parameters, and Zernike‑type static aberrations. The network learns the statistical relationship between reference‑mode and signal‑mode phase errors across a range of channel conditions and mode counts (N = 10, 30, 50).
During operation, only the reference pulses are measured with the MPLC; the trained network outputs estimated signal‑mode WFEs (Δ˜ϕ_mn,S). These estimates are used to drive a deformable mirror that shapes the wavefront of the RLO, thereby compensating the relative WFE before balanced homodyne detection of the quantum signal.
Simulation results show that the variance of the relative‑mode WFE is reduced by more than 70 % after correction, and that secure key rates, which are zero without correction, become positive (e.g., on the order of 10⁻³ bits per pulse) for the same channel parameters. The approach is robust to increasing mode numbers and to realistic levels of photon leakage and cross‑talk in the MPLC.
The authors discuss practical considerations such as MPLC insertion loss, deformable‑mirror response time, and neural‑network inference latency, and they outline future work including on‑orbit experiments, extension to multi‑wavelength or multi‑polarization schemes, and lightweight model deployment.
Overall, the study demonstrates that integrating advanced wavefront sensing with data‑driven phase estimation can substantially improve the performance and feasibility of satellite‑based CV‑QKD, paving the way for more resilient quantum‑secure global communication networks.