Deployed trusted-node quantum key distribution over 300 km with a multi-core fiber access link

This paper demonstrates the deployment of a 303 km trusted-node quantum key distribution link between Linköping University and Stockholm using commercial systems and superconducting detectors, successfully integrating QKD with co-propagating classical traffic and dynamic multi-core fiber switching while evaluating the impact of limited key rates on real-time encrypted image transmission.

The Gist

Imagine you want to send a top-secret message to a friend, but you are worried that a spy might be listening in. In the world of "Quantum Key Distribution" (QKD), you don't just send a secret code; you send the key to unlock the message using particles of light (photons). The magic of this system is that if a spy tries to peek at the key while it's traveling, the laws of physics say the key will change. The sender and receiver will immediately notice the change, know a spy is there, and throw that key away.

This paper describes a real-world experiment where scientists successfully built a "quantum highway" to test how well this technology works in a messy, busy, real-life environment.

Here is the story of their experiment, broken down into simple parts:

1. The Long Road Trip

The scientists wanted to connect two cities: Linköping and Stockholm. They didn't just build a new road; they used an existing "dark fiber" (a cable that is already buried underground but not currently carrying traffic) that stretched 270 kilometers.

To make the journey even more realistic, they added a 33-kilometer section at the end that acted like a busy city street. This section used a special type of cable called a Multi-Core Fiber (MCF). Think of this like a single highway cable that actually contains seven separate lanes inside it.

2. The "Trusted" Stopover

Because the distance was too far for the light signals to travel all the way in one go without fading away, they set up a "Trusted Node" (a secure stopover station) in the middle of the trip, near a town called Nyköping.

• The Analogy: Imagine you are sending a secret letter from Linköping to Stockholm. You can't throw the letter that far, so you stop at Nyköping. You give the letter to a trusted guard there. The guard reads it, puts it in a new sealed envelope, and sends it the rest of the way to Stockholm. As long as you trust the guard at Nyköping, your secret is safe.

3. The Busy Highway Test

In a real city, a highway isn't just for one car; it's full of traffic. To test if their quantum "car" could handle a busy road, the scientists did something clever:

• They used Lane 1 and Lane 6 of their 7-lane cable to send the secret quantum keys.
• They used Lane 7 to send normal, fast internet data (Ethernet), like streaming video or downloading files.
• They filled Lanes 2 through 5 with "noise" (like static on a radio) to simulate a very crowded, messy cable.

They even switched lanes during the experiment! They would send the quantum key through Lane 1 for a while, then switch it to Lane 6, while moving the internet traffic to Lane 1. This proved that the system could dynamically re-route itself, just like a GPS rerouting you around traffic, without breaking the connection.

4. The Super-Sensitive Eyes

The biggest challenge was that the signal gets very weak after traveling 300 kilometers. Standard detectors (like regular night-vision goggles) weren't sensitive enough to see the faint light.

• The Solution: The scientists used Superconducting Nanowire Single-Photon Detectors (SNSPDs).
• The Analogy: If a standard detector is like a person trying to hear a whisper in a noisy room, these super-detectives are like a person with super-hearing who can hear a single drop of water hitting the floor from a mile away. This allowed them to keep the secret key flowing even through the long, lossy cables and the noisy environment.

5. The Result: A Working, Busy Network

The experiment ran for 92 hours.

• They successfully generated secret keys the whole time, even while switching lanes and dealing with the "noise" from the other lanes.
• They showed that the system could handle the "traffic jams" (noise) and still produce keys, though the speed slowed down a bit as the noise got louder.
• They also showed how the system manages a "buffer" (a waiting room for keys). If one part of the trip is fast and the other is slow, the system saves the extra keys in the waiting room so the final connection doesn't stop.

6. The "One-Time Pad" Test

Finally, they tested what this actually looks like for a user. They used the generated keys to encrypt images (sending pictures securely).

• The Challenge: Sometimes the key generator is slow (like a faucet dripping slowly). If you try to send a high-quality photo, you might run out of keys before the picture is finished, resulting in a blurry or incomplete image.
• The Finding: They found that using modern, smart compression (like JPEG AI) helped a lot. It's like packing a suitcase more efficiently; you can fit more of the "picture" into the limited space of the keys you have, ensuring the image arrives clearly even when the key supply is low.

Summary

In short, this paper proves that Quantum Key Distribution isn't just a lab experiment anymore. It works on real, long-distance cables, it can survive alongside normal internet traffic, it can switch lanes on the fly, and it can be made robust enough to send real-world data like images, provided you use smart compression to manage the limited supply of secret keys.

Technical Summary

Technical Summary: Deployed Trusted-Node QKD over 300 km with Multi-Core Fiber Access Link

Problem Statement
Quantum Key Distribution (QKD) offers information-theoretic security guaranteed by quantum mechanics, contrasting with classical methods (e.g., RSA, ECC) that rely on computational complexity assumptions vulnerable to future quantum algorithms. While QKD has advanced in distance and speed, its integration into existing, heterogeneous telecommunications infrastructure remains a significant hurdle. Real-world deployment faces challenges including long-distance fiber attenuation, coexistence with classical traffic, and the need for dynamic network reconfiguration. Furthermore, many previous demonstrations of QKD over Multi-Core Fibers (MCFs) have been confined to controlled laboratory environments, lacking validation under the noise and operational constraints of deployed networks.

Methodology
The authors demonstrate a long-distance, trusted-node QKD system deployed in southeastern Sweden, connecting Linköping University (LiU) to the National Quantum Communication Infrastructure in Sweden (NQCIS) hub at KTH Stockholm. The total link distance is 303 km, composed of:

1. 270 km of deployed Single-Mode Fiber (SMF): Rented from GlobalConnect, divided into two sub-links (LiU to Nyköping [110 km] and Nyköping to Stockholm [160 km]) with an intermediate trusted node.
2. 33 km of Multi-Core Fiber (MCF): A spooled seven-core MCF inserted at the LiU node to emulate a metropolitan access link.

Key Technical Components:

• Hardware: Commercial QKD systems (ThinkQuantum QuKy EDU Pro) were customized to interface with external Superconducting Nanowire Single-Photon Detectors (SNSPDs). This modification was critical to overcome the high attenuation of the 303 km link, as standard internal InGaAs detectors lacked sufficient efficiency.
• Trusted Node Architecture: The intermediate node in Nyköping acts as a trusted relay. Key Management Systems (KMS) at each end generate keys for the physical sub-links ($K_{LiU-NYK}$ and $K_{NYK-KTH}$) and combine them to establish a secure end-to-end key ($K_{LiU-KTH}$).
• Spatial Division Multiplexing (SDM): The MCF access link utilizes active switching. The QKD channel and a 10 Gbit/s classical Ethernet channel are dynamically swapped between two specific MCF cores (Core 1 and Core 6).
• Noise Simulation: To simulate a realistic, "populated" fiber environment, broadband optical noise (50 nm bandwidth at 1550 nm) generated by Light Emitting Diodes (LEDs) was injected into the remaining four MCF cores (2–5). This created crosstalk noise on the QKD channel.
• Operational Testing: The system operated for 92 hours, actively switching the QKD and Ethernet channels between cores while varying the LED driving current (0 to 45 mA) to modulate crosstalk levels.

Key Results

• Detector Performance: The use of external SNSPDs significantly outperformed internal InGaAs detectors. For the 110 km sub-link, the SNSPDs achieved a Secret Key Rate (SKR) of $4.75 \pm 0.71$ kbit/s with a Quantum Bit Error Rate (QBER) of $0.5 \pm 0.2\%$, compared to $0.16 \pm 0.02$ kbit/s and $2.1 \pm 0.4\%$ QBER for InGaAs.
• Long-Distance Operation: The system successfully generated keys over the full 303 km link. The mean SKR was $87.2 \pm 53.4$ bit/s for the LiU-NYK sub-link and $24.6 \pm 10.3$ bit/s for the NYK-KTH sub-link. The end-to-end key rate was constrained by the lower-performing sub-link (NYK-KTH).
• Dynamic Switching and Resilience: The system successfully demonstrated active switching of the QKD channel between MCF cores. Upon switching, the system autonomously reinitialized, handling polarization rotation and transmission interruptions, with re-synchronization times ranging from tens of seconds to minutes.
• Crosstalk Tolerance: The QKD link remained operational despite increasing crosstalk from the injected LED noise. However, as expected, the SKR degraded as the LED current (and thus noise power) increased.
• Buffer Management: The KMS buffers effectively managed the fluctuating SKRs of the two sub-links. Surplus keys generated by the faster LiU-NYK link were stored and consumed when that link's rate dropped, smoothing the end-to-end key accumulation.
• Application-Level Impact: Using the generated keys for One-Time Pad (OTP) encryption, the authors tested image transmission under limited key budgets. They found that image fidelity was highly dependent on the available key budget and the compression algorithm used. Modern deep-learning-based compression (JPEG AI) significantly improved transmission efficiency compared to standard JPEG 2000 under these constrained conditions.

Significance and Claims
The paper claims to demonstrate the feasibility of integrating commercial QKD systems into demanding, dynamically reconfigurable fiber infrastructure relevant to future hybrid quantum-classical networks. Specifically, it highlights:

1. Real-World Validation: Moving beyond laboratory settings to a deployed, long-distance link with active traffic and noise simulation.
2. Hardware Adaptation: The necessity of external SNSPDs to support long-distance links with additional components (switches, MCFs) that introduce high loss margins.
3. Dynamic Network Integration: The successful operation of QKD with active core switching and co-propagating classical data, proving that QKD can coexist with standard telecom traffic in MCFs.
4. Application Constraints: The demonstration that limited and time-varying QKD throughput imposes strict constraints on encryption applications, necessitating advanced compression techniques to maintain data fidelity.

The authors conclude that while QKD is a promising solution for future-proof security, its practical deployment requires careful management of physical layer losses, noise, and application-level adaptation to variable key generation rates.