Specialists from the engineering-quantum laboratory SFB Lab and the Center for Quantum Technologies at Moscow State University have created a method to protect quantum key distribution systems from detector blinding attacks. The solution not only identifies the attack but also allows the system to retain the secure portion of the key instead of discarding the entire key material.
In a classic blinding attack, an adversary directs intense optical radiation at the single-photon detectors, temporarily blinding them. The attacker then injects carefully crafted light pulses that force the receiver to register chosen bits. When executed correctly, the attack reveals the full key without triggering conventional error-rate or detection alarms.
Existing countermeasures can detect such interference, yet they typically require the entire key to be discarded after an alarm. The new Russian method operates more selectively. The receiver randomly and slightly changes the quantum efficiency of its detectors at unpredictable moments. Because the attacker does not know the current efficiency state, no single pulse energy can reliably control the detector across all possible settings; the required energies differ by more than a factor of ten.
As a result, the attacker inevitably produces detectable anomalies such as erroneous clicks or double detections. Statistical analysis of these events enables the system to calculate how many bits may have been under adversarial control. Only those bits are removed during the privacy-amplification stage, while the remaining key material continues to be used.
The approach was tested on a production single-photon detector and does not require modifications to the optical layout. It is therefore suitable for a broad class of commercially available devices. The invention is protected by Russian patent RU 2 856 208 C1, which covers implementations for both single-detector and dual-detector receivers. The research results have been accepted for publication in the journal Physical Review Applied.
