Timing Mismatch Attack on Signal and Decoy States
The Timing Mismatch Attack exploits differences in emission timing between signal and decoy states generated by gain-switched laser diodes in QKD systems. This timing variation, arising from modulation of the laser’s pump current, allows the threat actor (Eve) to distinguish between signal and decoy states, enabling a more effective Photon Number Splitting (PNS) attack.
The same mismatch arises by a second route, between the separate laser diodes of a transmitter that gives each state its own emitter, and there it has been measured on an operational satellite link [Miller2025]. Either way the outcome is the same. The decoy-state method exists to tell the transmitter (Alice) and the receiver (Bob) how much of their key a multi-photon split could have leaked, and a threat actor who can label each pulse as signal or decoy makes that estimate wrong. The key the two parties keep is then partly known to her, and nothing in the error rate or the count rate they monitor shows it [Huang2018].
| [Huang2018] | A. Huang, S. Sun, Z. Liu, and V. Makarov. "Quantum key distribution with distinguishable decoy states" In: Phys. Rev. A 98, 01233. (2018) 10.1103/PhysRevA.98.012330. |
| [Miller2025] | Alexander Miller. "Micius, the world's first quantum communication satellite, was hackable". (2025) arXiv:2505.06532. |
| [Nauerth2009] | S. Nauerth, M. F¨urst, T. Schmitt-Manderbach, H. Weier, and H. Weinfurter. "Information leakage via side channels in freespace BB84 quantum cryptography" In: New J. Phys. 11, 065001. (2009) 10.1088/1367-2630/11/6/065001. |
| [Stehr2026] | U. Stehr, C. Dach, M. A. Hein. "Wideband Controllable Delay Line for the Synchronization of RF Pulses in Quantum Key Distribution Sources" (2026) 10.1109/GeMiC71240.2026.11516381. |
| [Tamaki2016] | K. Tamaki, M. Curty, and M. Lucamarini. "Decoy-state quantum key distribution with a leaky source" In: New Journal of Physics 18 (2016), p. 065008. (2016) 10.1088/1367-2630/18/6/065008. |
4e9565a7-47d0-4271-bfa0-33881aa5ac60
- Quantum Communication
Technique properties
Operational properties
| Description | Countermeasure |
|---|---|
|
State preparation monitoring checks the transmitter's own output, and the pulse timing is one of the parameters it checks. The signal-decoy timing mismatch is exactly that observable: [Huang2018] measured it at the output of the source, and [Miller2025] reconstructed it for the Micius transmitter. |
CM-0016 State Preparation Monitoring |
|
Keeping the laser at a constant drive and carving the intensity with a separate external modulator decouples the two jobs in hardware, so that "the pulse generation and intensity modulation in this type of source are physically decoupled and performed by separate devices. As long as there is no electrical crosstalk between the laser diode driver and intensity modulator driver, no correlation is expected" [Huang2018]. That conditional is part of the control: crosstalk between the two drivers would reintroduce the coupling the architecture removes. The fix is measured rather than asserted, on the same bench and with the same instruments as the defect. Arrival-time distinguishability falls from for the pump-current source to for the externally modulated one, which is the difference between no secure key at any distance and 83 km under the refined proof [Huang2018]. This is the strongest single answer to the pump-current route of the attack. |
CM-0063 Employment of External Intensity Modulator |
|
The primary does not merely recommend considering this attack in a proof. It adapts the leaky-source framework of [Tamaki2016] to a leak the transmitter emits unaided, characterises the imperfection by the trace distance between the states conditioned on the intensity setting, and derives two closed-form bounds on the single-photon yield and error rate that hold for distinguishability in any degree of freedom, not only time [Huang2018]. That generality is what makes the control worth applying here rather than waiting for each new axis to be measured: the same accounting absorbs the spectral chirp the primary could not measure. The control prices the leak rather than removing it, and on the source actually measured the price is everything. At a trace distance of 0.4005 "The secure key rate is zero under either key rate estimation. This shows that the modulation imperfection can make the system insecure" [Huang2018]. A proof that absorbs this imperfection therefore tells an operator the truth about a transmitter he may already be running, which is its value, rather than making that transmitter safe. |
CM-0014 Full Device Imperfections Incorporation in Scientific Security Proof |
| Description | Category | Impact Type |
|---|---|---|
|
Electronics and Analog Components is affected as a knock-on, and it is where the defence lives. What sets the emission time is the electrical drive: on the single-diode route the pump current, which reaches population inversion sooner at higher amplitude so that "the higher the current, the less time" it takes [Huang2018]; on the multi-diode route the RF path that carries each control pulse, where the residual misalignment comes from cable and microstrip length differences on the PCB and LTCC plus FPGA routing [Stehr2026]. The requirement is stated as an analog-electronics requirement, under 2 ps between channels at a 1 GHz clock, which "cannot be achieved by careful radio frequency line design alone" [Stehr2026]. |
L2 — Electronics & Analog Components | Secondary Effect |
|
Post-Processing and Cryptographic is where the damage is realised, and this is the mapping that explains why the attack has no error-rate signature. The decoy comparison is a post-processing computation; once the intensity class leaks it prices multi-photon leakage wrongly, so that "Alice and Bob's key remains partially insecure after post-processing" [Huang2018]. The consequence is reported as a key rate at this layer, zero at any distance for the source measured [Huang2018] and zero on the real Micius downlink [Miller2025]. Every proof-side countermeasure on this row acts here rather than on the optics. |
L5 — Post-Processing & Cryptographic | Secondary Effect |
|
Physical Layer (Quantum and Optical) is the primary target, because the leak happens in the emitted light itself. The signal pulse "emits earlier than the decoy state with high probability" and has a secondary peak, and the cause is semiconductor turn-on delay and relaxation oscillations in the laser's own optical output [Huang2018]. On the second route the same quantity is the firing offset between eight separate optical emitters, measured in orbit at 312 +/- 20 ps against a 200 ps pulse width [Miller2025]. The threat actor's own action is optical too, since she intercepts and re-forwards photons on the span. |
L1 — Physical Layer (Quantum & Optical) | Primary Target |
| Relation type | Description | Technique |
|---|---|---|
| Refines |
This technique refines the Photon-Number-Splitting Attack by supplying it with an input the plain attack does not have: a per-pulse label saying whether that pulse was signal or decoy. Plain photon-number splitting is answered by the decoy-state method, because comparing the yields of two intensities bounds how much a splitter can have taken. Once the intensities are distinguishable by arrival time that bound stops holding, since the threat actor tunes separate observation windows to match the announced gain per intensity and so "the attack will remain unnoticed" [Huang2018]. |
A-0044 Photon-Number-Splitting (PNS) Attack |
|