Skip to main content
QUICK REVIEW

[Paper Review] Modulation leakage vulnerability in continuous-variable quantum key distribution

Nitin Jain, Ivan Derkach|arXiv (Cornell University)|Mar 26, 2021
Quantum Information and CryptographyComputer Science38 references29 citations
TL;DR

This paper identifies a modulation leakage vulnerability in continuous-variable quantum key distribution (CVQKD) systems using in-phase and quadrature (IQ) modulators for single-sideband encoding. Due to limited sideband suppression, secret key information leaks through the suppressed sideband, significantly reducing key rates or breaching security—especially under direct reconciliation. Countermeasures using trusted noise can mitigate this risk, particularly in reverse reconciliation schemes.

ABSTRACT

Flaws in the process of modulation, or encoding of key bits in the quadratures of the electromagnetic light field, can make continuous-variable quantum key distribution systems susceptible to leakage of secret information. Here, we report such a modulation leakage vulnerability in a system that uses an optical in-phase and quadrature modulator to implement a single sideband encoding scheme. The leakage arises from the limited suppression of a quantum-information-carrying sideband during modulation. Based on the results from a proof-of-concept experiment, we theoretically analyse the impact of this vulnerability. Our results indicate that the leakage reduces the range over which a positive secret key can be obtained, and can even lead to a security breach if not properly taken into account. We also study the effectiveness of additional trusted noise as a countermeasure to this vulnerability.

Motivation & Objective

  • . To identify and analyze a previously unreported vulnerability in CVQKD systems arising from imperfect sideband suppression in IQ modulators.
  • . To quantify the impact of modulation leakage on secret key rates under both direct and reverse reconciliation protocols.
  • . To evaluate the effectiveness of trusted noise injection as a countermeasure against leakage-induced security degradation.
  • . To provide a theoretical and experimental framework for assessing leakage effects in practical CVQKD implementations.

Proposed method

  • . Develops a theoretical model of an IQ modulator based on Mach-Zehnder interferometers with phase modulators, using Jacobi-Anger expansions to describe sideband generation.
  • . Models the suppressed sideband as an unintended information carrier, treating it as a quantum side channel accessible to an eavesdropper (Eve).
  • . Conducts a proof-of-concept experiment using an IQ modulator with variable sideband suppression, measuring leakage effects on key rates.
  • . Applies quantum information theory to compute Holevo information and secret key rates, incorporating leakage into the security analysis.
  • . Evaluates reconciliation techniques (direct and reverse) under varying leakage levels and quantifies the resulting key fraction degradation.
  • . Investigates the role of trusted noise (preparation and detection) in mitigating leakage effects, distinguishing between noise injected in signal, leakage, or detection paths.

Experimental results

Research questions

  • RQ1. How does imperfect sideband suppression in an IQ modulator lead to information leakage in CVQKD systems using single-sideband encoding?
  • RQ2. What is the quantitative impact of modulation leakage on secret key rates under direct and reverse reconciliation protocols?
  • RQ3. Can trusted noise injection effectively reduce the security penalty caused by leakage, and if so, under what conditions?
  • RQ4. How does the choice of reconciliation technique (DR vs. RR) influence the system’s resilience to leakage?
  • RQ5. What are the practical implications of this vulnerability for future integrated photonic CVQKD systems?

Key findings

  • . Modulation leakage increases from 0.063 to 0.19 bits/symbol under reverse reconciliation and from 0.15 to 0.99 bits/symbol under direct reconciliation as sideband suppression degrades from 24 dB to 4 dB.
  • . The direct reconciliation protocol becomes insecure at moderate leakage levels (|ρ| > 3 dB), while reverse reconciliation remains functional but with significantly reduced key rates.
  • . The maximum tolerable additional channel loss drops when leakage is present, meaning Alice and Bob may mistakenly assume security where none exists if they ignore leakage.
  • . Trusted preparation noise in the leakage mode (εL) can improve key rates, especially benefiting direct reconciliation under high leakage.
  • . Detection noise (εD) is only effective in reverse reconciliation and cannot compensate for leakage in direct reconciliation.
  • . The vulnerability is more pronounced in photonic integrated circuits due to tighter tolerances and higher likelihood of RF mismatch, making this a critical concern for future CVQKD deployments.

Better researchstarts right now

From reading papers to final review, dramatically reduce your research time.

No credit card · Free plan available

This review was created by AI and reviewed by human editors.