Detector Efficiency Mismatch Attack

System

The Detector Efficiency Mismatch Attack exploits the differential response times and sensitivities of the detectors in a quantum key distribution system. By manipulating the timing of the pulses, the threat actor (Eve) can ensure that only one of the two detectors is active at any given moment, allowing her to control the detection events and thereby gain knowledge of the key material without detection.

Timing is only the best known of several routes. The same trick works on any parameter Eve can set that the two detectors respond to differently: the polarisation of the incoming light, its spatial mode in a free-space link, its wavelength, a tampered calibration routine, and, since 2026, the pulse energy itself [Zaitsev2026]. In the strongest case analysed, a total mismatch, Eve gets a complete copy of the key while the receiver (Bob) sees the click rate and the bit statistics he expects [Makarov2006]. The one attack run end to end leaked part of the key [Zhao2008].

Literature 24
[Bochkov2019] M. Bochkov and A. Trushechkin. "Security of quantum key distribution with detection-efciency mismatch in the single-photon case: Tight bounds" In: Phys. Rev. A 99, 032308. (2019) 10.1103/PhysRevA.99.032308.
[BSI2023] BSI. "Implementation Attacks against QKD Systems". (2023) https://www.bsi.bund.de/SharedDocs/Downloads/EN/BSI/Publications/Studies/QKD-Systems/QKD-Systems.pdf.
[Chaiwongkhot2019] P. Chaiwongkhot, K. B. Kuntz, Y. Zhang, A. Huang, J.-P. Bourgoin, S. Sajeed, N. Lütkenhaus, T. Jennewein, and V. Makarov. "Eavesdropper’s ability to attack a free-space quantum-key-distribution receiver in atmospheric turbulence" In: Phys. Rev. A 99, 062315. (2019) 10.1103/PhysRevA.99.062315.
[Fatin2021] M. Fatin and S. Sajeed. "Generalized efciency mismatch attack to bypass the detection-scrambling countermeasure" In: Opt. Express 29, 16073-16086. (2021) 10.1364/OE.419338.
[Fei2015] Y. Fei, M. Gao, W. Wang, C. Li, and Z. Ma. "Practical attacks on decoy-state quantum key systems with detector efficiency mismatch" In: Phys. Rev. A 91, 052305. (2015) 10.1103/PhysRevA.91.052305.
[Fei2018a] Y. Fei, X. Meng, M. Gao, Z. Ma, and H. Wang. "Exploiting wavelength-dependent beam splitter to attack the calibration of practical quantum key distribution systems" In: Optik 170, 368–375. (2018) 10.1016/j.ijleo.2018.05.089.
[Fung2009] C. Fung, K. Tamaki, B. Qi, H. Lo, and X. Ma. "Security proof of quantum key distribution with detection efficiency mismatch" In: Quantum Information and Computation 9, 131–165. (2009) 10.26421/QIC9.1-2-8.
[Jain2011] N. Jain, C. Wittmann, L. Lydersen, C. Wiechers, D. Elser, C. Marquardt, V. Makarov, and G. Leuchs. "Device Calibration Impacts Security of Quantum Key Distribution" In: Phys. Rev. Lett. 107, 110501. (2011) 10.1103/PhysRevLett.107.110501.
[Jain2014] N. Jain, E. Anisimova, I. Khan, V. Makarov, C. Marquardt, and G. Leuchs. "Trojan-horse attacks threaten the security of practical quantum cryptography" In: New J. Phys., vol. 16, no. 12, p. 123030. (2014) 10.1088/1367-2630/16/12/123030.
[Liu2014] Q. Liu, A. Lamas-Linares, C. Kurtsiefer, J. Skaar, V. Makarov, and I. Gerhardt. "A universal setup for active control of a single-photon detector" In: Rev. Sci. Instrum. 85, p. 013108. (2014) 10.1063/1.4854615.
[Lydersen2011d] L. Lydersen, V. Makarov, and J. Skaar. "Secure gated detection scheme for quantum cryptography" In: Phys. Rev. A 83, 032306. (2011) 10.1103/PhysRevA.83.032306.
[Makarov2005] V. Makarov and D. Hjelme. "Faked states attack on quantum cryptosystem" In: Journal of Modern Optics, 52(5), 691–705. (2005) 10.1080/09500340410001730986.
[Makarov2006] V. Makarov, A. Anisimov, and J. Skaar. "Effects of detector efficiency mismatch on security of quantum cryptosystems" In: Phys. Rev. A 74, 022313. (2006) 10.1103/PhysRevA.74.022313.
[Makarov2008] V. Makarov and J. Skaar. "Faked states attack using detector efficiency mismatch on SARG04, phase-time, DPSK, and Ekert protocols" In: Quant. Inf. Comp. 8, 0622–0635. (2008) 10.26421/QIC8.6-7-4.
[Qi2007] B. Qi, C. Fung, H. Lo, and X. Ma. "Time-shift attack in practical quantum cryptosystems" In: Quant. Inf. Comp. 7, 073–082. (2007) 10.26421/QIC7.1-2-3.
[Rau2015] M. Rau, T. Vogl, G. Corrielli, G. Vest, L. Fuchs, S. Nauerth, and H. Weinfurter. "Spatial Mode Side Channels in Free-Space QKD Implementations" In: IEEE Journal of Selected Topics in Quantum Electronics 21, 187–191. (2015) 10.1109/JSTQE.2014.2372008.
[Sajeed2015a] S. Sajeed, P. Chaiwongkhot, J.-P. Bourgoin, T. Jennewein, N. Lütkenhaus, and V. Makarov. "Security loophole in free-space quantum key distribution due to spatial-mode detector-efficiency mismatch" In: Phys. Rev. A 91, 062301. (2015) 10.1103/PhysRevA.91.062301.
[Silva2015] T. da Silva, G. do Amaral, G. Xavier, G. Temporo, and J. von der Weid. "Safeguarding Quantum Key Distribution Through Detection Randomization" In: IEEE Journal of Selected Topics in Quantum Electronics, vol. 21, no. 3, pp. 159-167. (2015) 10.1109/JSTQE.2014.2361793.
[Taylor2026] B. J. Taylor, P. R. Smith, J. F. Dynes, R. I. Woodward, M. Lucamarini, R. M. Stevenson, A. J. Shields. "Practical countermeasure against attacks exploiting detection-efficiency mismatch in quantum key distribution" (2026) 10.1103/7zhl-5vv1.
[Trushechkin2022] A. Trushechkin. "Security of quantum key distribution with detection- efficiency mismatch in the multiphoton case" In: Quantum 6, 771. (2022) 10.22331/q-2022-07-22-771.
[Wei2019] K. Wei, W. Zhang, Y. Tang, L. You, and F. Xu. "Implementation security of quantum key distribution due to polarization-dependent efficiency mismatch" In: Phys. Rev. A 100, 022325. (2019) 10.1103/PhysRevA.100.022325.
[Zaitsev2026] K. Zaitsev, V. Bizin, D. Kuzmin, V. Makarov. "Energy-time attack on detectors in quantum key distribution" (2026) 10.1103/nmdx-c94q. arXiv:2603.07538.
[Zhang2021a] Y. Zhang, P. Coles, A. Winick, J. Lin, and N. Lutkenhaus. "Security proof of practical quantum key distribution with detection-efficiency mismatch" In: Phys. Rev. Research 3, 013076. (2021) 10.1103/PhysRevResearch.3.013076.
[Zhao2008] Y. Zhao, C. Fung, B. Qi, C. Chen, and H. Lo. "Quantum hacking: Experimental demonstration of time-shift attack against practical quantum-key-distribution systems" In: Phys. Rev. A 78, 042333. (2008) 10.1103/PhysRevA.78.042333.
QID: A-0069
UUID: 75e582a5-ecbe-4b48-97ad-9f3b206d805b
Tier: T0
Type: Quantum
Domain(s):
  • Quantum Communication
Tactic: Execution
Created: 2024-04-09
Updated: 2026-11-18
Maturity: Experimental
Technique properties

Attack type: Key extraction
Attack category: Degree-of-freedom coupling attack
Target party: Receiver
Target function: Quantum layer, System control and management
Quantum channel: Fibre optic, Free space
Security layers: L1 — Physical Layer (Quantum & Optical) L2 — Electronics & Analog Components L4 — System Control & Orchestration
Maturity: Experimental
Operational properties

Technique → Countermeasures 7
Description Countermeasure

Random jitter in the detector synchronisation is Makarov2006's second protection measure against this technique, in its own words: "introduce intentional random jitter in the detector synchronization to" smear the curves and lower the mismatch. It does not remove the mismatch, it reduces the effective ratio between the two detectors' efficiencies, which is the quantity that decides whether the attack is available at all [Makarov2006].

CM-0043 Detector with Random Jitter

Specifying the worst-case detector-efficiency mismatch during the design, manufacture and quality assurance of the detectors and the system is the countermeasure that [Makarov2008] proposes for this attack, because finite manufacturing precision always leaves some mismatch. The study finds the mismatch and does not remove it: the worst-case value has to be carried into the security proof and the privacy amplification [Makarov2008].

CM-0086 Prior Hardware Component Study

Bit-mapped gating randomises the mapping between detection events and bit values, optically and in software, and confines legitimate detection to a narrow window. Against this technique it is the same family of defence as Makarov2006's third protection measure: "implement active protection by checking timing of incoming pulses at Bob", done "by registering the time of avalanche onset within the window, or with additional detectors" [Makarov2006,Lydersen2011d].

CM-0039 Bit-Mapped Gating

Gating-window discrimination is Makarov2006's own third protection measure against this technique: "implement active protection by checking timing of incoming pulses at Bob", done "by registering the time of avalanche onset within the window, or with additional detectors". It catches the classic time-shift route, where the threat actor's pulse arrives outside the normal window [Makarov2006].

CM-0061 Detector's Dead Time Management

Active polarisation-state scrambling inside the receiver closes the polarisation axis of this technique, where a polarising beam splitter combined with polarisation-dependent detector efficiency lets the threat actor favour one detector [Makarov2005,Wei2019]. It is the per-axis analogue of what four-state Bob achieves for every axis at once in single-mode fibre, cancelling mismatch "regardless of the auxiliary degree of freedom responsible" [Taylor2026]. It leaves the temporal, spatial-mode, calibration and pulse-energy routes open, and Makarov2006 already shows how far those reach, since faked states "can be constructed for energy-time encoding and differential phase shift keying QKD schemes" [Makarov2006].

CM-0046 Active Polarisation State Scrambling

A spectral filter at the receiver input closes one degree of freedom only, the wavelength axis, by rejecting the light the threat actor would use to reach a wavelength-dependent efficiency difference between the two detectors. It leaves the temporal, polarisation, spatial-mode and pulse-energy axes of this technique untouched, and Makarov2006 already shows the family reaches well beyond one axis, since faked states "can be constructed for energy-time encoding and differential phase shift keying QKD schemes" [Makarov2006,Fei2018a].

CM-0013 Employment of Optical Filters

Against this technique the security proof does not stop the attack, it prices it in. Makarov2006 states the requirement precisely: the worst-case estimate of the mismatch parameter δ\delta, and not the QBER, "should be used to determine the required amount of privacy amplification", and this "requires the amount of privacy amplification to be corrected in most practical quantum cryptosystems that use four-state protocols". The proof literature for this specific case is on the target row [Fung2009,Trushechkin2022,Zhang2021a,Bochkov2019].

CM-0014 Full Device Imperfections Incorporation in Scientific Security Proof
Technique → Security Layers 3
Description Category Impact Type

System Control and Orchestration. The calibration route reaches this layer through the timing calibration, as in A-0117. Secondary rather than primary, because the key is still extracted at the optical and analog layers; L4 is the route in [Zaitsev2026].

L4 — System Control & Orchestration Secondary Effect

Electronics & Analog Components — the temporal sub-variant exploits the gated-detector electronics (sinusoidal-gate timing, dead-time recovery, avalanche transient). The energy–time effect [Zaitsev2026] is squarely an analog-electronics phenomenon.

L2 — Electronics & Analog Components Secondary Effect

Physical Layer (Quantum & Optical) - the attack is delivered as optical pulses parameterised on a degree-of-freedom axis and lands on the receiver's (Bob's) basis-resolving optics and SPAD bank. Makarov2006 makes the generality explicit: the control parameter "could be not necessarily time but, e.g., polarization or wavelength" [Makarov2006]. The axes that land here at L1 are the optical ones - time, polarisation, wavelength, spatial mode and pulse energy. Pulse energy is the newest of them, since "the click occurs progressively earlier" by more than 2 ns as energy rises across a 50 dB range on a detector whose whole gate period is 3.2 ns [Zaitsev2026]; its analog-electronics half is recorded separately at L2, and the calibration route is an L4 concern carried by the L4 mapping rather than here.

L1 — Physical Layer (Quantum & Optical) Primary Target
See Also
A-0067 Wavelength-Selected Photon-Number-Splitting Attack
A-0068 Double-Click Attack
A-0070 Blinding Attack with Faked States
A-0074 Detector Control via After-Gate Pulses Attack
A-0075 Detector Control by Exploiting Superlinearity