[논문 리뷰] Towards Device-Independent Quantum Key Distribution with Photonic Devices

Corentin Lanore, Xavier Valcarce외 3명|arXiv (Cornell University)|2026. 01. 20.

Quantum Information and Cryptography인용 수 0

한 줄 요약

이 논문은 기기 간 독립 QKD의 실현 가능성을 기계 학습으로 식별된 광학 회로를 사용해 평가하고, 실용적 키 생성을 현실적인 손실과 노이즈 하에서 달성할 수 있음을 보이기 위해 효율적인 SDP 기반 엔트로피 한계와 유한 크기 분석을 도입합니다.

ABSTRACT

Quantum Key Distribution (QKD) protocols enable two distant parties to communicate with information-theoretically proven secrecy. However, these protocols are generally vulnerable to potential mismatches between the physical modeling and the implementation of their quantum operations, thereby opening opportunities for side channel attacks. Device-Independent (DI) QKD addresses this problem by reducing the degree of device modeling to a black-box setting. The stronger security obtained in this way comes at the cost of a reduced noise tolerance, rendering experimental demonstrations more challenging: so far, only one experiment based on trapped ions was able to successfully generate a secret key. Photonic platforms have however long been preferred for QKD thanks to their suitability to optical fiber transmission, high repetition rates, readily available hardware, and potential for circuit integration. In this work, we assess the feasibility of DIQKD on a photonic circuit recently identified by machine learning techniques. For this, we introduce an efficient converging hierarchy of semi-definite programs (SDP) to bound the conditional von Neumann entropy and develop a finite-statistics analysis that takes into account full outcome statistics. Our analysis shows that the proposed optical circuit is sufficiently resistant to noise to make an experimental realization realistic.

연구 동기 및 목표

• Motivate DIQKD as a secure paradigm robust to device imperfections.
• Assess a photonic optical circuit identified by machine learning for DIQKD feasibility.
• Develop an efficient SDP hierarchy to bound conditional von Neumann entropy.
• Extend finite-size security analysis to full outcome statistics for DIQKD.
• Demonstrate realistic experimental parameters that enable secret-key generation.

제안 방법

• Model the DIQKD protocol with one classical and one quantum channel and rounds of state distribution.
• Use a converging SDP block hierarchy to bound the conditional von Neumann entropy H(A1|E) based on full statistics P(a,b|x,y).
• Apply entropy accumulation theorem (EAT) with a general I-score to perform finite-size security analysis.
• Perform asymptotic key rate analysis using rDW and r_block,ell,m for CHSH-based and full-statistics-based bounds.
• Optimize circuit parameters (g, α, β, p) to maximize key rate; compare computational cost of r0 vs r_block,ell,m.

실험 결과

연구 질문

• RQ1 Can a photonic circuit identified for DIQKD sustain Bell violation under realistic loss and noise to enable key generation?
• RQ2 How does using full outcome statistics P(a,b|x,y) compare to CHSH-based estimates for key rate in DIQKD?
• RQ3 Can an efficient SDP (block hierarchy) bound H(A1|E) tightly enough for practical finite-size analyses?
• RQ4 What finite-size resources (rounds, efficiency) are required to produce a positive secret key with realistic detector efficiency?
• RQ5 What experimental parameter regime (η, g, α, β, p) yields feasible DIQKD within hours at MHz repetition rates?

주요 결과

• The optical circuit studied shows resistance to noise sufficient for potential experimental realization of DIQKD.
• Using full statistics to compute the key rate reduces the required efficiency threshold by about 4% at the 10^-5 key-rate level compared to CHSH-based bounds, and can yield up to an order of magnitude higher key rate than CHSH-based estimates at higher efficiency.
• Finite-size analysis with full statistics significantly lowers the minimum number of rounds needed to extract a secret key, potentially enabling feasible experiments.
• For a repetition rate of 1 MHz, an efficiency of 87.5% suffices to produce a secret key within approximately 8 hours (n ≈ 3×10^10 rounds).
• Optimal parameters cited include squeezing Tg = 0.249 (or 2.163 dB), p = 0.042, and specific displacement values (α1, α2, β0, β1, β2) achieving positive key rates.
• The block SDP-based bound H(A1|E)Block,ell,m provides tighter bounds than prior approaches with linear memory scaling in m, improving feasibility assessments.