Source-linked AI summary
Real-world two-photon interference and proof-of-principle quantum key distribution immune to detector attacks
Allison Rubenok, Joshua A. Slater, Philip Chan, Itzel Lucio-Martinez, Wolfgang Tittel
TL;DR
Detector vulnerabilities motivate measurement-device-independent QKD, which removes attack strategies tied to imperfections in the measurement apparatus. The paper implements and analyzes a proof-of-principle system, while identifying remaining assumptions and unimplemented requirements that limit the demonstration’s scope.
Problem
QKD implementations face side channels from detector control and other imperfections, while security proofs for realistic MDI-QKD implementations still require further development.
Method
The paper implements MDI-QKD using stabilized two-photon interference and a three-intensity decoy-state analysis to bound single-photon quantities.
Results
The experiment demonstrates previously undemonstrated MDI-QKD requirements, including Bell state measurement over deployed fiber and positive secret-key-rate measurements over multiple configurations.
Takeaways & Limitations
MDI-QKD represents a significant step toward closing the gap between theoretical security proofs and experimentally viable implementations by eliminating measurement-device attack strategies.
Takeaways & Limitations
The proof-of-principle system did not implement phase randomization, random modulation, error correction, or finite-key-size effects required by the security assumptions.
Abstract
from arXiv · showhide
Several vulnerabilities of single photon detectors have recently been exploited to compromise the security of quantum key distribution (QKD) systems. In this letter we report the first proof-of-principle implementation of a new quantum key distribution protocol that is immune to any such attack. More precisely, we demonstrated this new approach to QKD in the laboratory over more than 80 km of spooled fiber, as well as across different locations within the city of Calgary. The robustness of our fibre-based implementation, together with the enhanced level of security offered by the protocol, confirms QKD as a realistic technology for safeguarding secrets in transmission. Furthermore, our demonstration establishes the feasibility of controlled two-photon interference in a real-world environment, and thereby removes a remaining obstacle to realizing future applications of quantum communication, such as quantum repeaters and, more generally, quantum networks.
ENSURING INDISTINGUISHABILITY
The implementation used three stabilization systems to preserve photon indistinguishability for Bell state measurements in a real-world environment.
- ENSURING INDISTINGUISHABILITY: Three stabilization systems addressed polarization, photon arrival time, and laser frequency.Polarization was stabilized automatically, while timing and frequency were adjusted manually.
- ENSURING INDISTINGUISHABILITY: Every 10 s, an additional laser and polarization controllers used stabilization light to adjust the polarization at Alice and Bob.The stabilization interval included 0.5 s of disabled data collection.
- ENSURING INDISTINGUISHABILITY: A frequency shifter corrected Alice’s and Bob’s laser-frequency difference when the x-key error rate increased significantly.In the worst case, adjustments were required every 30 minutes.
- ENSURING INDISTINGUISHABILITY: A master clock and periodic arrival-time measurements maintained the qubit arrival-time difference under 30 ps.Charlie measured the difference roughly every minute and communicated adjustments to Alice and Bob.
DECOY-STATE ANALYSIS
The decoy-state analysis bounds single-photon gains and error rates from measurable multi-intensity data, enabling a lower bound on the secure secret key rate.
- DECOY-STATE ANALYSIS: The secret key rate in MDI-QKD is calculated from single-photon gains, error rates, binary entropy, and error-correction efficiency.The analysis uses h2(X) for binary entropy and f for error-correction efficiency relative to Shannon’s theorem.
- DECOY-STATE ANALYSIS: Gains are projection probabilities onto |ψ−⟩, while error rates are erroneous-to-total projection ratios in the x- or z-basis.Subscript 11 denotes Alice and Bob both sending single photons; µσ denotes their mean photon numbers.
- DECOY-STATE ANALYSIS: The three-intensity decoy method uses signal, decoy, and vacuum intensities to derive a lower bound for Qx_11 and an upper bound for ex_11.These bounds are used to calculate a lower bound for the secure secret key rate.
- DECOY-STATE ANALYSIS: The decoy-state equations estimate single-photon quantities from Poissonian photon-number probabilities and observed vacuum-related gains.The method uses bounds on Qx_11 and ex_11, with analogous equations for the z-basis.
- DECOY-STATE ANALYSIS: The relevant single-photon gains and error rates are uniquely determined through measurable gains and error rates.The experiments selected µd = 0.05 to obtain statistically significant data in a reasonable amount of time.
SECURE KEY DISTRIBUTION USING MDI-QKD
The paper identifies assumptions and implementation gaps relevant to secure MDI-QKD while demonstrating progress toward experimentally viable security. Several proof-of-principle requirements remain unmet or require further study.
- SECURE KEY DISTRIBUTION USING MDI-QKD: MDI-QKD eliminates attack strategies related to imperfections in the measurement apparatus, but secure deployment still requires hardware and proofs addressing remaining imperfections.The authors emphasize that systems must be vetted against implementation attacks and that further theoretical developments are needed.
- SECURE KEY DISTRIBUTION USING MDI-QKD: The implementation assumes private laboratories, phase-randomized attenuated pulses, and sufficiently close qubit states forming two maximally conjugate bases.These assumptions concern shielding, source preparation, and state-generation quality.
- SECURE KEY DISTRIBUTION USING MDI-QKD: The pulses were not phase-randomized because subsequent pulses carved from a long-coherence laser beam remained coherent.Adding a phase modulator to randomize each qubit’s global phase was identified as the solution.
- SECURE KEY DISTRIBUTION USING MDI-QKD: The demonstrated system did not implement random choices of mean photon number or encoded state during the proof-of-principle experiment.Instead, it held each choice for several minutes before changing it.
- SECURE KEY DISTRIBUTION USING MDI-QKD: The measured state-overlap deviations were 0.074 across different bases and 0.013 for different states in the same basis, which current proofs judged insufficient for secure key distribution.The authors describe these proof bounds as conservative and note that technological improvements could reduce deviations to around 1 part in 1000.
- SECURE KEY DISTRIBUTION USING MDI-QKD: The experiment did not implement error correction or finite-key-size effects, instead estimating error-correction efficiency and assuming infinitely long data collection.Finite-key effects had only been investigated using an overly conservative approach for MDI-QKD.
- SECURE KEY DISTRIBUTION USING MDI-QKD: The work demonstrates previously undemonstrated MDI-QKD requirements while leaving several topics for future investigation and experimental improvement.The authors specifically highlight Bell state measurement over deployed fiber and optimization of decoy-state analysis.
DISCUSSION OF ERROR RATES ex,z
The ideal x- and z-basis error rates differ because a |ψ−⟩ detection can arise from same-person photons in the x basis, but not in the z basis. Thus, x-basis detections can yield uncorrelated key bits, whereas z-basis detections produce identical bits after Bob’s flip.
- z-basis: In the z basis, a |ψ−⟩ projection requires orthogonal photons from Alice and Bob, so Bob’s bit flip makes their key bits identical.Two photons from the same person would occupy the same state and cannot produce the required projection.
- x-basis: In the x basis, same-person photons can also produce the |ψ−⟩ detection pattern, unlike in the z basis.This additional event class changes the error-rate behavior between the bases.
- x-basis: Same-person x-basis detections provide no correlation between Alice’s and Bob’s prepared states, leading to uncorrelated key bits.Because all detected photons may come from only one party, the detection does not reveal a relation between their states.
- Error-rate mechanism: The ideal µσ error rate is governed by the relative probabilities of one photon arriving from each party versus two photons arriving from the same party.For Poissonian attenuated laser pulses, the analysis assumes equal probabilities of photons arriving from either party.