Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

Relative Wavefront Errors in Continuous-Variable Quantum Communication

Nathan K. Long, John Wallis, Alex Frost, Benjamin P. Dix-Matthews, Sascha W. Schediwy, Kenneth J. Grant, Robert Malaney

TL;DR

This work addresses a key assumption in continuous-variable QKD over atmospheric channels: that the signal and local oscillator share identical wavefront distortions. The authors test this by performing a laboratory experiment with a shared laser generating a weak signal and a strong LO, polarization-multiplexed and frequency-shifted, transmitted through a turbulent free-space channel and analyzed with a multi-plane light converter to decompose fields into Hermite-Gaussian modes. They define and measure relative wavefront errors via $ΔP_{mn} = (P_{mn,s}/P_s) - (P_{mn,lo}/P_{lo})$ across the first eight modes, finding nonzero values for all turbulence levels and increasing variance with stronger turbulence, with a KS test confirming statistical significance at 99.99% confidence. The results challenge standard CV-QKD deployment assumptions and motivate wavefront-correction approaches, including machine-learning-based methods to reduce excess noise and improve key rates in atmospheric links, with particular relevance to future terrestrial-satellite quantum networks.

Abstract

When undertaking continuous-variable quantum key distribution (CV-QKD) across atmospheric channels, strong classical local oscillators (LOs) are often polarization-multiplexed with the weak quantum signals for coherent measurement at the receiver. Although the wavefronts of the quantum signal and LO are often assumed to experience the same distortion across channels, previous theoretical work has shown that they can experience differential distortions, resulting in relative wavefront errors (WFEs). Such errors have previously been shown to limit CV-QKD performance, in some cases leading to zero secure key rates. In this work, for the first time, we provide strong experimental evidence that relative WFEs are present in some circumstances and that standard assumptions in CV-QKD deployments may need to be revisited. In addition, we demonstrate how turbulence can affect the detailed form of the relative WFEs, thereby indicating that long-range links like terrestrial-satellite channels are likely impacted more than short-range terrestrial-only channels.

Relative Wavefront Errors in Continuous-Variable Quantum Communication

TL;DR

This work addresses a key assumption in continuous-variable QKD over atmospheric channels: that the signal and local oscillator share identical wavefront distortions. The authors test this by performing a laboratory experiment with a shared laser generating a weak signal and a strong LO, polarization-multiplexed and frequency-shifted, transmitted through a turbulent free-space channel and analyzed with a multi-plane light converter to decompose fields into Hermite-Gaussian modes. They define and measure relative wavefront errors via across the first eight modes, finding nonzero values for all turbulence levels and increasing variance with stronger turbulence, with a KS test confirming statistical significance at 99.99% confidence. The results challenge standard CV-QKD deployment assumptions and motivate wavefront-correction approaches, including machine-learning-based methods to reduce excess noise and improve key rates in atmospheric links, with particular relevance to future terrestrial-satellite quantum networks.

Abstract

When undertaking continuous-variable quantum key distribution (CV-QKD) across atmospheric channels, strong classical local oscillators (LOs) are often polarization-multiplexed with the weak quantum signals for coherent measurement at the receiver. Although the wavefronts of the quantum signal and LO are often assumed to experience the same distortion across channels, previous theoretical work has shown that they can experience differential distortions, resulting in relative wavefront errors (WFEs). Such errors have previously been shown to limit CV-QKD performance, in some cases leading to zero secure key rates. In this work, for the first time, we provide strong experimental evidence that relative WFEs are present in some circumstances and that standard assumptions in CV-QKD deployments may need to be revisited. In addition, we demonstrate how turbulence can affect the detailed form of the relative WFEs, thereby indicating that long-range links like terrestrial-satellite channels are likely impacted more than short-range terrestrial-only channels.

Paper Structure

This paper contains 3 sections, 3 figures.

Table of Contents

  1. Introduction
  2. Experimental Results
  3. Conclusion

Figures (3)

  • Figure 1: Experimental setup. LS is a laser source, NPB is a non-polarizing beamsplitter, Pol is a polarizer, PBS is a polarizing beamsplitter, Ch is the channel, NW is the nichrome wire, HWP is a half-waveplate, MPLC is a multi-plane light converter, $S$ is the signal path and $LO$ is the LO path. $P_{mn,s}$ and $P_{mn,lo}$ are the signal and LO powers.
  • Figure 2: Time versus ${\Delta P_{mn}}$ for turbulence Cases 0-3 (where C represents Case).
  • Figure 3: HG mode versus ${\Delta P_{mn}}$ variance for turbulence Cases 0-3 (where C represents Case).