Architecture and Hardware News

Secure Quantum Communications

Data locking experiments provide stepping stones to a possible future in quantum cryptography.
  1. Article
  2. Author
  3. Figures
Secure Quantum Communications, illustration

Two independent experiments published on the same day last August (August 12, 2016) have demonstrated the potential for quantum mechanics to improve the efficiency of secure data communications.

The experiments mark a departure from the current mainstream proposal for quantum communications: quantum key distribution (QKD). First demonstrated in 1989, QKD was designed to provide classical communications channels with a more secure method for delivering keys through the use of quantum mechanics.

“QKD uses a quantum channel that is able to transmit quantum states to establish a secure link between a sender and an intended receiver. But the information is actually transmitted over a classical channel, such as a telephone line,” says Daniel Lum, postgraduate researcher at the University of Rochester, and the lead author on the paper in Physical Review A that describes one of the two new experiments on what its proponents call quantum data locking (QDL).

Yang Liu, lead author of the other QDL paper in Physical Review A and a researcher at the Shanghai branch of the Hefei National Laboratory for Physical Sciences at the Microscale of the University of Science and Technology of China, adds: “QKD is a mature technology today, both from a theoretical and from a practical view. The implementation and performance of QKD is good enough for commercial products. It is an important primitive for encryption. The QDL protocol, on the other hand, is relatively new.”

David DiVincenzo and colleagues from IBM Watson Research Center and the University of Gdansk proposed the basis for QDL in 2004. Although it has attracted far less academic and industrial attention than QKD, it has developed over the past decade into a family of theoretical encryption techniques.

“QDL shows a phenomenon that is unique to quantum information theory; it is not possible in classical information theory. In application, we have demonstrated the QDL protocol is able to encrypt a message and send it over tens of kilometers of fiber,” Liu claims.

Although it is much younger and faces a number of challenges, the attraction of QDL over QKD is that it is potentially far more efficient in terms of how much information can be encrypted for each bit of key than any system that relies on classical communication.

To provide provably secure communication, protocols in use today need to obey a theory developed by Claude Shannon in the 1940s. The encryption key, which must be generated randomly, needs to be the same length or greater than the information content of the message itself. Shannon’s theory provided support for the one-time pad developed in the late 19th century, in which sender and receiver agree to use a common key—originally taking the form of characters written on a pad of paper—only once. Once the message had been received and decoded, the key was to be discarded.

QKD provides the means for two parties, Alice and Bob, to agree on a secret key without risk of it being obtained by an eavesdropper. The protocol takes advantage of the way in which an attempt to determine one part of the quantum state of a particle disturbs the others. It makes it impossible to completely determine the quantum state of a photon or particle and, as a result, copy it.

Under QKD, when taking measurements of a sequence of photons they exchange, Alice and Bob agree to randomly swap between two different types of measurement of the quantum state and then compare the results. Eve can intercept the photon, perform her own measurements, and attempt to copy the photon and pass it on to Bob. They cannot, however, determine the state of the other property, and the new photon will be forwarded with a state that probably does not match Alice’s original.

Without an eavesdropper like Eve present, Alice and Bob’s measurements will match approximately half the time because of their random switching between properties. With an eavesdropper present, the error rate rises significantly, because of the 50% probability for each photon that the eavesdropper has picked the wrong measurement to perform. But if enough of Alice’s and Bob’s measurements agree, the received pattern becomes a shared private key that can be used to encrypt messages on another channel, which can use traditional classical coding techniques.

One problem for QKD is the limit on communication speed caused by the nature of the protocol itself, combined with the effects of noise and interference in the quantum channel. Stefano Pirandola, a researcher at the University of York, says QKD protocols based on encoding pairs of properties into ‘qubits’ tend to deliver very low key-update rates. One way to boost the update rate is to use continuous-variable properties such as the quadrature operators of the coherent light transmissions from lasers. These quadrature operators “play the same role that position and momentum play for a particle such as an electron,” he says.

The need to use lengthy keys for message delivery still leaves QKD-based systems facing a potential bottleneck. QDL can harness the difficulties eavesdroppers have in intercepting quantum channels to send the data bits themselves and use exponentially shorter keys than those needed for Shannon’s one-time pad system.

To employ QDL, Alice and Bob first agree on a shared key, which could be generated using QKD. That key selects a set of codewords that determine the sequence of properties to be measured and their contents. Each codeword calls for a different sequence of measurements on the quantum states. As with QKD, Eve can only access a fraction of the complete message, even with access to unlimited computing power.

QKD makes it impossible to completely determine the state of a photon or particle and, as a result, copy it.

“I think QDL is an interesting approach that relies on the realistic assumption that today, an eavesdropper cannot do everything and can only access quantum memories with limited lifetimes,” says Pirandola.

Liu explains, “We performed two experiments using our setup. The first was to show the original data-locking idea. The protocol locks half of the message using a 1-bit key. With a key length of one, the maximum information the eavesdropper may obtain is half of the message Alice sent.

“The second experiment was towards more practical schemes, limiting Eve’s information to an arbitrarily small amount using logarithm-length keys.”

Using free-space transmission rather than fiber allowed Lum’s team to explore higher dimensions of encoding based on more complex combinations of quantum properties to allow the transmission of error-correction bits along with the message encoded into the photon’s state. However, there is a trade-off inherent in the use of error correction; the redundancy it introduces makes it easier for an adversary to decrypt messages. As a result, a higher key ratio is needed to guarantee security and successful communication over noisy channels.

Liu and Lum both stress the experiments they performed were proof-of-concept demonstrations. Some of the theoretical requirements for QDL are not possible to realize in practice. For example, the experiments by both teams used the same technique as that used for QKD to generate pairs of photons. However, spontaneous parametric down-conversion is a random process that can create more than two daughter photons, with uncertain timing. Neither is desirable for QDL.

“Many reviewers pointed out that our experiment was not stringent enough to be considered truly secure; we cannot guarantee that we limited the accessible information of an eavesdropper to an arbitrarily small amount because of optical losses, efficiencies, and the inability to transmit one photon on demand,” Lum says.

“The QDL scheme is still in its infancy; it shows new physics and reveals possibilities for new applications. There is still more science to be performed.”

“The main weakness of the QDL, I believe, is in quantum-channel losses. To reliably transmit messages via quantum states is a challenging problem and any unpredictable changes to the quantum states in transmission will corrupt the data.”

One obstacle to both QKD and QDL is the question of distance. Experiments have demonstrated the ability to transmit photons that retain entanglement over several hundred kilometers in free space, and 150km in fiber. “For long-range communication, preserving the quantum state over a long-range quantum channel is a formidable challenge; many believe it simply isn’t practical,” Lum concedes.

The choice of quantum encoding also will limit transmission distance: “Continuous-variable systems are limited to metropolitan distances because of technological issues, but they completely out-perform qubit-based protocols in terms of achievable rates for QKD,” says Pirandola.

Evidence for the practicality of long-distance quantum communications may come from experimental satellites launched this year. Researchers from the SpooQy Lab at the Center for Quantum Technologies in Singapore planned to launch a satellite payload into low Earth orbit to perform quantum-communication experiments in 2014, but the launcher exploded shortly after liftoff. The researchers aim to have a replacement in space in the autumn, but they will follow Chinese researchers who had their Quantum Experiments at Space Scale (QUESS) satellite successfully inserted into a solar-synchronous orbit 600km above sea level in August. The Chinese satellite will relay quantum transmissions over thousands of miles between ground stations in China and Europe.

As with much of the research into communications, because of the protocol’s relative maturity, the satellite projects will focus on QKD issues such as preserving entanglement over large separations. But if the experiments are successful, they should demonstrate that QDL and other quantum protocols that may be developed have a practical future in communications security.

“QDL is not going to outperform QKD anytime soon, which is why many experts in quantum cryptography do not regard the QDL demonstrations as high impact. We acknowledge this and present our experiment as a stepping stone to a possible future in quantum cryptography,” Lum says.

Liu adds, “The QDL scheme is still in its infancy; it shows new physics and it reveals possibilities for new applications. There is still more science to be performed.”

*  Further Reading

Liu, Y., et al
Experimental Quantum Data Locking. Physical Review A, 94, 020301 (2016). Preprint:

Lum, D., et al
A Quantum Enigma Machine: Experimentally Demonstrating Quantum Data Locking. Physical Review A, 94, 022315 (2016) Preprint:

Wilde, M.M.
Quantum Information Theory, Second Edition. Cambridge University Press (2016). Preprint:

Cheng, C., Chandrasekara, R., Chuan, T.Y., and Ling, A.
Space-Qualified Nanosatellite Electronics Platform for Photon Pair Experiments. IEEE/OSA Journal of Lightwave Technology, Vol. 33, 4799 (2015)

Back to Top

Back to Top


UF1 Figure. Experimental set-up for client-server reference frame independent quantum key distribution (rfiQKD). Source: P. Zhang, et al. arXiv:1308.3436 [quant-ph]

Back to top

Join the Discussion (0)

Become a Member or Sign In to Post a Comment

The Latest from CACM

Shape the Future of Computing

ACM encourages its members to take a direct hand in shaping the future of the association. There are more ways than ever to get involved.

Get Involved

Communications of the ACM (CACM) is now a fully Open Access publication.

By opening CACM to the world, we hope to increase engagement among the broader computer science community and encourage non-members to discover the rich resources ACM has to offer.

Learn More