Large-scale quantum communication networks with integrated photonics
Summary
This Nature paper demonstrates a proof-of-principle, lab-scale integrated-photonics twin-field quantum key distribution (TF-QKD) network — the ‘Weiming Quantum Chip-Network’. The system uses a Si3N4 microcomb at the central server and 20 monolithic InP QKD transmitter chips as clients. The microcomb supplies many ultralow-noise, phase-coherent optical lines for wavelength-division-multiplexed (WDM) TF-QKD. The authors ran pairwise TF-QKD across ten WDM channels (20 client chips), reaching a maximum client-to-client span of 370 km and an aggregate networking capability N·L/2 = 3,700 km, and showed secure key rates that surpass the PLOB repeaterless bound at 370 km.
Key Points
- Demonstrated an integrated TF-QKD ‘Weiming’ network using a Si3N4 microcomb server and 20 InP client transmitter chips.
- Microcomb provides Hz-level linewidth, ultralow-noise, phase-coherent comb lines suitable for large-scale WDM TF-QKD.
- Pairwise TF-QKD implemented over ten wavelength channels (20 chips), achieving per-pair distances up to 370 km and total capability N·L/2 = 3,700 km.
- At 370 km the network’s secure key rates exceeded the PLOB repeaterless bound across all channels.
- Client chips monolithically integrate lasers, phase and intensity modulators and VOAs; fabrication yield and uniformity are high (97.5% modulator yield reported).
- Laser injection-locking to comb lines regenerates Hz-level linewidths on-chip (locked slave lasers ≈60 Hz linewidth), enabling stable interference for TF-QKD.
- Measured QBER in the phase-sensitive X-basis was low (~2.9–4.2% across distances tested), supporting reliable key extraction with finite-key analysis.
- Remaining issues: imperfect phase randomisation, potential injection-locking loopholes, and further server-side integration (SNSPDs, frequency shifters) needed for field deployment.
Content summary
The authors build a star-topology TF-QKD network where a central integrated microcomb distributes many phase-coherent seed lines through backbone fibres to geographically separated client chips. Each InP client chip contains two DBR lasers and multiple electro-optic modulators to prepare decoy/signal states and perform dual-λ phase tracking. Injection locking aligns local DBR lasers to comb lines, locally regenerating ultralow-noise light for encoding.
The experiment implemented a sending-or-not-sending TF-QKD protocol across ten WDM channels, sequentially performing pairwise key exchanges while keeping all channels transmitted in parallel (to emulate crosstalk/noise). They ran links with upstream spans of 102+102 km (204 km) and 185+185 km (370 km). The network achieved low QBERs and secure-key rates consistent with theory; notably the 370 km configuration surpassed the PLOB bound on repeaterless transmission.
Fabrication metrics are emphasised: 20 InP chips were randomly selected from a 3-inch wafer with high device yield and consistent modulator performance; Si3N4 microresonators show mean intrinsic Q ≈ 20 million and stable dark-pulse microcomb spectra. The paper reports phase-tracking methods (dual-λ) and practical control cycles adapted to drift rates for different link lengths.
Context and relevance
This work addresses two major hurdles for scaling QKD networks simultaneously in user number (N) and distance (L): mass-manufacturable low-noise transmitters and centrally shared ultrastable frequency references. Integrated microcombs plus monolithic InP transmitters create a path to WDM, many-user TF-QKD where expensive phase stabilisation resources are centralised and shared. The demonstration ties recent long-distance TF-QKD protocol advances to real device-manufacturability and system engineering, making it directly relevant to researchers and engineers pushing quantum-safe communications toward telecom integration and practical deployments. It also shows a route to extend links further (towards km-scale repeater-like ranges and eventually 1,000 km) if combined with lower-loss fibres, better SNSPDs and protocol optimisation.
Why should I read this?
Short answer: if you care about real-world quantum-secure networks, this is one of the first papers that actually stitches together chip-scale devices, comb-based phase references and advanced TF protocols into a multi-user, long-haul demo. It’s a neat reality-check — the parts are manufacturable, the noise budgets are workable and the approach scales. Read it for the device numbers, the phase-tracking tricks and the roadmap to scaling WDM TF-QKD.