New technical note AGE-PN-001: reproducible benchmark for WDM stability See evidence
Exploratory line · No validated results yet

Physical integrity in quantum links. We measure what security proofs take for granted.

The security of quantum key distribution (QKD) is proven for devices that meet certain conditions. One is that pulses of different intensity or basis are indistinguishable except in what they encode. Real equipment does not always comply: differences in pulse shape or timing have been shown to let an attacker tell them apart.

What AGE contributes

Fidelity bounds with uncertainty, not a “looks fine”.

Recent security proofs already tolerate imperfect sources, provided you know how far they deviate. That number has to be measured. AGE researches how to measure it well.

01

Measured temporal distinguishability

We estimate how similar the temporal modes of pulses that should be identical really are, from noisy measured traces. The output is a fidelity bound with its confidence interval, not a “looks fine”. It is the same overlap mathematics used by our temporal-mode simulators.

02

Monitoring that logs before it corrects

Our drift-estimation software separates two modes. In normal operation, it compensates. In monitoring mode, it detects, logs and preserves evidence without silent correction, because automatic compensation could hide a manipulation.

Stated limits

What AGE is not.

The line is published as exploratory research: nothing described here is experimentally validated or a product.

Out of scope

  • Not a cryptographic system, and it replaces no protocol.
  • It does not produce security proofs or certify equipment.
  • A detected deviation does not prove an attack. It may be drift, a fault or tampering, and telling them apart requires an explicit threat model.
First deliverable

AGE-PN-002.

A technical note with open code that estimates the distinguishability between signal and decoy pulses from their temporal shape, and shows how thermal drift degrades it. Same format as AGE-PN-001: MIT code, tests, data and published negative results.

AGE-PN-002 · deliverablesin preparation
  • age-qkd-integrity package (MIT) with tests.
  • A 4–6 page technical note with three figures.
  • A sealed-forecast protocol adapted to pulse traces.
  • Zenodo publication with DOI.
Collaboration

We are looking for QKD groups.

If your group characterises QKD sources or receivers and has temporal pulse traces, we want to test the estimate against your measurements, using the same sealed-forecast protocol we use in photonics. Shared authorship, and your data stays under your control.

Sources cited on this pagearXiv · DOI · ISO
  • Huang, Sun, Liu, Makarov — «Decoy-state QKD with imperfect source». Phys. Rev. A 98, 012330 (2018). arXiv:1711.00597
  • Zhao, Fung, Qi, Chen, Lo — «Experimental demonstration of time-shift attack». Phys. Rev. A 78, 042333 (2008). arXiv:0704.3253
  • Currás-Lorenzo, Pereira, Kato, Curty, Tamaki — «A security framework for QKD with imperfect sources». Optica Quantum 3, 525 (2025). arXiv:2305.05930
  • ISO/IEC 23837-1:2023 · ISO/IEC 23837-2:2023 — security requirements and evaluation methods for QKD. ISO/IEC 23837-1 · ISO/IEC 23837-2