Nanjing Team Demonstrates Asynchronous Quantum Conferencing
A recent study from Nanjing University suggests that adding a fourth participant to a secure quantum conference call can be more efficient than simply adding a third point-to-point link. Led by Xiao-Song Ma, the team conducted what they term asynchronous measurement-device-independent quantum cryptographic conferencing (AMDI QCC). This setup allowed three users to share a secret key without the need to trust any central “box.”
One Quantum Channel, Many Secure Listeners
Indeed, traditional multi-user quantum conferencing relies on multiphoton coincidence events, which means multiple detectors must register a photon at precisely the same moment. The frequency of these events decreases rapidly as fiber loss increases and more users join in. The asynchronous protocol, theoretically proposed by Zeng-Bing Chen and Hua-Lei Yin’s group, bypasses this bottleneck by matching single-photon detections recorded at different times. This means detections don’t need to occur simultaneously; they later combine through post-processing.
The New Conference-Key Result Explained
The reported mechanism works as follows: Three users independently send optical pulses towards a shared GHZ measurement station. Here, a fiber-based multipath interferometer allows these signals to interfere and be detected as single photons. The station publicly announces its results. Users then combine this announcement with their own encoding, pair compatible single-photon events, map keys, and perform classical post-processing to arrive at a shared conference key. The measurement-device-independent design ensures that even if an attacker controls the detectors, they cannot compromise the key, removing the need to trust the central node.
A key achievement is the improved loss tolerance. The team successfully generated secure keys with approximately 59.6 dB of total system loss, a significant increase from about 21.5 dB in the same group’s previous work on polarization-encoded MDI conferencing (TQI). At this highest loss, the reported secure key rate was about 4.470 × 10⁻⁹ bits per pulse.
The Significance of the Efficiency Claim
In a group call, pairwise QKD would necessitate each participant having separate private calls, or routing everyone through a single operator who must be trusted not to eavesdrop. A conference-key protocol, on the other hand, functions as an actual conference call, where a single shared session provides everyone with the same secret simultaneously. This design logic is the source of the excitement.
The scaling detail is particularly noteworthy. In an ideal N-user network, the loss-scaling exponent stops increasing with the number of users. While conventional MDI conferencing performance degrades with each added user, the asynchronous version keeps the exponent tied to the single-user channel transmittance (η). The loss penalty no longer compounds across the group.
What This Research Doesn’t Yet Show
Three users do not constitute a network. Real-world deployment will require more participants, longer fiber distances, and a finite-key security proof that holds up under genuine noise conditions. None of these aspects are fully clarified by the available evidence. The system still requires phase estimation, although the team ingeniously used time-multiplexed reference pulses and a fast Fourier transform to compensate for laser frequency drift without physical phase locking – a valuable simplification that doesn’t eliminate the need for the interferometer.
There is no claim of quantum advantage here, nor should there be, as this is a communications primitive, not a computation. No cryptographically relevant quantum computer currently exists, which means this work isn’t looking to defend against a present-day threat. Furthermore, conference key agreement is not a NIST post-quantum standard. For standardized algorithmic security, one would look to FIPS 203, 204, and 205, which operate on a different layer. QKD and QCC secure key exchange through physics; they do not replace mathematical cryptography.
The Broader Quantum Security Landscape
This result is situated within a broader push in quantum infrastructure, which supports its direction without validating the experiment itself. WISeKey, SEALSQ, and the Canton of Jura have signed an MoU to establish a Swiss post-quantum semiconductor and cybersecurity center. Altera and Riverlane have partnered on real-time quantum error-correction control hardware. Fujitsu has open-sourced a quantum application package to enhance the software layer.
In summary, this research presents a genuine leap in loss tolerance, a clever phase estimation technique, and a compelling scaling argument on paper. However, it is not yet a quantum conference service available for purchase.