Access Official Technical PDF

Share Summary:LinkedInEmail

Quantum key distribution (QKD) is designed to protect key exchange through the principles of quantum mechanics. But the security of a real system also depends on how its hardware is built and shielded. Research on a commercial QKD platform and a prototype quantum random number generator (QRNG) demonstrated that light entering through ventilation openings can reach sensitive optical components. This light injection attack illustrates why quantum key distribution security must account for unintended paths into device enclosures, not only the intended optical channel.

Why optical access matters in quantum devices

Many QKD systems use very weak optical signals, often operating at the single-photon level or with carefully controlled weak coherent pulses. Their security assumptions depend in part on controlling the light entering and leaving the system. If an external source can introduce additional photons through an unexpected route, the device may behave differently from the model used to assess its security.

Ventilation openings are a potential route because quantum equipment may contain optical fibers, lasers, modulators, and detectors that require thermal management. If an opening provides a line of sight to an exposed fiber or other optical component, light can couple into the device. Fiber losses that normally allow some light to escape can also permit light from outside to enter. In the demonstrated setup, coupling was associated with a combination of Rayleigh scattering and fiber-bend effects.

This is distinct from an attack through the normal communications fiber. The intended channel is an expected point of entry and can be monitored or protected with measures such as filters. An optical path through a ventilation opening may bypass those controls; once injected photons enter an internal fiber, they can travel alongside legitimate signals.

How a light injection attack can affect QKD

The researchers tested a plug-and-play QKD platform in which a long fiber spool served as a delay line. Light introduced through a ventilation opening coupled into the spool and could reach Alice’s phase modulator. When the modulator encoded a phase, injected photons passing through it could acquire information about that setting. After reflecting from a Faraday mirror and traversing the optical path back toward the communications channel, some light could leave the device.

This resembles a Trojan-horse attack, in which an adversary sends light into a system to learn information about its internal state. The important difference is the entry route: in this case, light entered through an unintended opening rather than the monitored communications input. The experiment established that photons could be injected, but the coupling was inefficient and the impact depended on the amount of light reaching the relevant components.

The security concern is a mismatch between the photon number assumed by the protocol and the actual light leaving the device. If Alice and Bob underestimate that output, they may overestimate how much secret key can safely be extracted. The study’s analysis found that injected light could reduce the secure operating range under particular modeled conditions. This does not mean every QKD device is vulnerable in the same way; exposure depends on its optical layout, enclosure, protocol, and protective measures.

QRNGs can also be influenced by external light

Quantum random number generators rely on measurements of quantum events to produce bit sequences. In the tested prototype, photons were directed to one of two detectors, and the detector outcomes affected whether the output signal changed between logic states. Additional light entering through a ventilation opening could increase detections in one detector and bias the resulting sequence.

A strongly biased output is a clear warning that a generator is not producing the intended distribution. More generally, even less obvious bias can matter because cryptographic processes depend on suitable randomness. Statistical tests can help identify some problems, but they should not be treated as a substitute for preventing unintended optical access or for monitoring a generator’s operation.

Practical design measures for quantum hardware

The findings point to enclosure design as part of the security architecture, alongside protocol and software controls. Engineers and system integrators evaluating quantum optical equipment can consider the following measures:

  • Block direct optical paths. Cover sensitive fibers, connectors, pigtails, and detectors with opaque material so external light cannot reach them.
  • Review ventilation geometry. Openings should support cooling without providing a clear line of sight to optical components. Bends or substantial obstructions can help block probing paths, but should be assessed for their ability to prevent light coupling as well.
  • Inspect component placement. Fiber spools and exposed optical assemblies near vents or enclosure edges deserve particular attention, including during maintenance and integration.
  • Use monitoring as a supporting control. Input monitoring and filtering remain useful on the intended optical channel, but may not detect light entering through an unanticipated route.
  • Include optical injection in verification. Security reviews and tests should consider external illumination of the enclosure, rather than focusing solely on conventional physical access or electromagnetic emissions.

Monitoring and debiasing can provide additional safeguards in QRNGs, but the study notes that such controls may not address every refined attack. Preventing light from reaching sensitive components is therefore a central design objective.

Conclusion: treat the enclosure as a security boundary

Light injection through ventilation openings shows that the physical construction of a quantum device can affect its security assumptions. In the demonstrated cases, external light could reach a QKD delay line and influence a QRNG’s output. The practical lesson for equipment designers, plant teams, and technical decision-makers is to assess optical access wherever sensitive components are exposed. Secure enclosure design, blocked line of sight, and targeted testing can help ensure that ventilation does not become an unintended signal path.

Access Official Technical PDF

About The Author