Researchers Set New Distance Record for Quantum Key Distribution

Jul 21, 2009 By Lisa Zyga feature
Scientists have demonstrated a new QKD prototype that can distribute quantum keys over a distance of 250 km in the lab. Credit: Wikimedia.

(PhysOrg.com) -- Quantum key distribution (QKD) could be the next commercial success of quantum physics, and a recent study has taken the field a step closer to this reality. Researchers from the University of Geneva in Switzerland and Corning Incorporated in New York have demonstrated a new QKD prototype that can distribute quantum keys over a distance of 250 km in the lab, improving upon the previous record of 200 km. The scientists hope that the achievement will lead to the goal of distributing quantum keys over intercity distances of 300 km in the near future.

As the researchers explained, the purpose of QKD schemes is to distribute a secret quantum key between two distant locations with security relying on the laws of . The idea of QKD was first proposed in 1984, and in 1992, scientists could distribute quantum keys over 32 cm, while the technology has improved from there. Despite these advances, the scientists say, the main challenge is still to achieve higher bit rates over longer distances.

To reach their new record of 250 km, the scientists made three significant improvements to their QKD technique. First, they developed a coherent one way (COW) protocol tailored specifically for quantum communication over optical fiber networks. In addition, they used an improved superconducting single-photon detector to decrease noise, as well as ultra low loss fibers made by Corning to minimize channel loss and improve the distribution rate.

By making these improvements, the physicists could distribute quantum keys in the lab at a rate of 15 bits per second over 250 km of optical fiber, or 6,000 bits per second over 100 km, with low error rates. The system is also fully automated, and can run for hours without human intervention.

In the COW protocol, Alice, the transmitter, sends a pair of pulses, one empty and one non-empty (containing a mean photon number of 0.5). The bits are encoded in the pair of pulses, with the bit value define by the position of the non-empty pulse: first position = 0 and second position = 1. Bob, the receiver, uses a detector to distinguish the pulses.

For true , Alice and Bob also verify the coherence of the pulses. Bob randomly selects a small fraction of pulses, not used as data, to send to an interferometer, which measures the coherence between adjacent qubits. Due to this security measure, an eavesdropper could not perform “photon number splitting attacks,” such as removing or blocking photons, without disturbing the system and being detected. This heightened security means that quantum key distribution could have applications in quantum cryptography, explained Hugo Zbinden of the University of Geneva.

“As the name says, QKD enables the distribution of a secret key,” Zbinden told PhysOrg.com. “This key can be used for the encryption and decryption of secret messages. … Since QKD allows for a secure and convenient key exchange, it enables changing the used keys frequently and therefore improves the security of the system.”

As the researchers noted, the high performance of the COW QKD prototype shown here depended on a team of experts, including theoretical and experimental physicists, telecom engineers, and electronic and software specialists. They hope that this diverse set of skills will continue to improve the performance of quantum key distribution for useful applications.

“In order to improve commercial QKD, we need convenient, fast and low noise photon detectors,” Zbinden said. “The superconducting detectors are not very practical. Then, going to higher repetition rates, high speed electronics and logic become mandatory, and you need fast key distillation algorithms. Another issue is to combine QKD with classical links on the same fiber using WDM (wavelength division multiplexing). The development of a QKD system is therefore a multi-disciplinary task, requiring physicists as well as electronics, software and telecom engineers.”

More information: D. Stucki, N. Walenta, F. Vannel, R.T. Thew, N. Gisin, H. Zbinden, S. Gray, C.R. Towery, and S. Ten. “High rate, long-distance quantum key distribution over 250 km of ultra low loss fibres.” New Journal of Physics. 11 (2009) 075003.

Copyright 2009 PhysOrg.com.
All rights reserved. This material may not be published, broadcast, rewritten or redistributed in whole or part without the express written permission of PhysOrg.com.

Explore further: Duality principle is 'safe and sound': Researchers clear up apparent violation of wave-particle duality

add to favorites email to friend print save as pdf

Related Stories

Toshiba Plug Loophole in Security of Quantum Cryptography

Feb 20, 2007

Toshiba Research announced today that it has developed two new technologies to realize 'unconditionally secure' quantum key distribution (QKD). To achieve this, Toshiba has overcome a potential security loophole in current ...

Long-distance record -- 'Quantum keys' sent 200 kilometers

Jun 01, 2007

Particles of light serving as “quantum keys”—the latest in encryption technology—have been sent over a record-setting 200-kilometer fiber-optic link by researchers from the National Institute of Standards and Technology ...

Recommended for you

User comments : 4

Adjust slider to filter visible comments by rank

Display comments: newest first

holoman
1 / 5 (1) Jul 21, 2009
Nice, eventually quantum repeaters will
allow for any distance in fiber. But imagine
doing this with EMF versus Photons, now we
are talking about extending the transmission
even farther and by learning to entangle matter
the distance would be infinite.
PinkElephant
not rated yet Jul 22, 2009
So.... what prevents a "man in the middle" attack? Employing a "quantum repeater", to use holoman's terminology, an eavesdropper could splice itself into the fiber, and spoof the entire 2-way communication, while successfully intercepting all the traffic. Since all pulses in both directions are absorbed by the eavesdropper, and then sent out again by the same eavesdropper, they will maintain both their coherence and timing.

Am I missing something here?
RayCherry
not rated yet Jul 24, 2009
The man in the middle would immediately be detected, as the recieving station would receive a modified (time delayed) key, and the sending station would immediately change communication settings and encryption algorithm. Plus, the device playing man in the middle could be located within a narrow geographical channel between the sending and receiving stations.

No matter how powerful the silent eavesdropping repeater, the signal would register a regular delay at receiver as the signal drops from fiber to copper and then retransmitted in fiber. If the eaves dropper attempted a very difficult fibre 'T-juntion', the receiver would still report packet dropages druring the change-over and the signal strength would be altered as the 'T' cyphened off enough signal to read the transmission.

The delay, signal strength reduction, (and optical signature alteration) would identify a compromised communication channel. Very annoying, but at least the encrypted data stream can be halted with minimal delay (almost simultaneous to the hack attempt) and therefore the least loss of data.

Of course, Transmitters, Receivers and Repeaters all have software that can be targetted, and regretably successful attacks will take place. Protecting the physical network will generate as much work as protecting the data.
PinkElephant
not rated yet Jul 24, 2009
No matter how powerful the silent eavesdropping repeater, the signal would register a regular delay at receiver as the signal drops from fiber to copper and then retransmitted in fiber.


This "regular delay" would be measured in nanoseconds. Can the sender and receiver be so accurately clock-synchronized, that the receiver would actually detect the time differential outside of error tolerances? Let's say this 300-km cable warms up or cools down by some fraction of a degree; the transmission time thus fluctuates naturally by comparable amounts, I would think...

...the receiver would still report packet dropages druring the change-over...


That assumes continuous operation. However, since we're talking about key distribution, the actual "on time" would probably be rather sparse.

The delay, signal strength reduction, (and optical signature alteration) would identify a compromised communication channel.


Hm. I'm not so sure about the viability of detecting the delay. Signal strength reduction shouldn't be possible, since we're talking about single-photon signal: as long as you re-send one photon for every photon you receive, and make sure it has the same wavelength and polarization, then there shouldn't be any detectable strength reduction or optical signature alteration...