Faster quantum crypto demoed

By Eric Smalley, Technology Research News

There has been tremendous progress in quantum cryptography in recent years, and one system is already available commercially. But there's a long way to go before the technology matches its promise, and one of the biggest issues is coming up with devices that reliably generate and detect single photons at high speeds.

Using single photons isn't the only path to perfectly secure communications, however. Working out how to use only standard telecommunications gear to transmit cryptographic keys could dramatically improve quantum cryptography's paltry performance.

Cryptographic keys are numbers used to mathematically scramble and unscramble secret messages. The trick to using secret keys to encrypt messages is making sure they get to the intended recipient and no one else.

A group of researchers from the Institute of Optics in France and the Free University of Brussels (ULB) in Belgium has demonstrated a quantum cryptographic system that uses ordinary, weak laser pulses of several hundred photons each rather than single-photon pulses that entail special equipment. "Producing and detecting single photon pulses is not easy," said Phillippe Grangier, research director at the Institute of Optics. "Our scheme avoids these difficulties by using a completely different technique [that is] closer to standard optical telecommunications," he said.

The multi-photon method taps the nature of pulses instead of individual photons to guard against eavesdropping,

Initial results from the researchers' table-top prototype show that the multi-photon scheme should allow for faster information transmission than single-photon schemes over distances shorter than ten kilometers, said Grangier. In simulations of transmissions over short distances, the researchers were able to transmit secret keys as fast as 1.7 megabits, or millions bits, per second, according to Grangier. Today's single-photon quantum cryptographic systems transmit secret keys at speeds of a few hundred to a few thousand bits per second.

At the equivalent of just over 15 kilometers, the system transmitted data at about 75 kilobits per second, said Grangier. "Our system should allow much higher secret bit rates over short distances, but might be more sensitive to losses [and thus] may not work well over large distances," he said.

Single-photon systems use the polarizations of individual photons to represent each bit of a random string of 1s and 0s. Photons can be polarized in one of four paired orientations: horizontal and vertical and the two diagonals. The horizontal orientations can represent 0 and the vertical orientation can represent 1, for example.

The Heisenberg uncertainty principle explains why the system provides perfect security: a photon cannot be measured for both pairs of orientations. In order to eavesdrop on single photons, an eavesdropper -- Eve -- would have to replace the photons she intercepts, but because she would only be able to correctly measure half of them she would have to guess at the other half. The sender and receiver could compare a few of their photons, and if a quarter of them failed to match they would be tipped off to Eve's presence. They would discard those photons and try again until they got an unobserved string of photons. Over an ordinary communications line, Alice tells Bob which photons he measured correctly and they use those corresponding bits as a key.

The researchers' method uses weak laser pulses that contain a few hundred photons each, but taps the quantum nature of the pulses' amplitude and phase to secure the information they carry, said Gangier. Quantum physics describes the rules for individual atoms, photons and other particles. Lasers are quantum devices because their photons flow in lockstep, and the state of a laser pulse can only be measured according to quantum rules, including the Heisenberg uncertainty principle, said Grangier.

Instead of using the polarization of single photons to represent the bits of a key, a sender -- Alice -- would randomly choose an amplitude and phase for each weak laser pulse she sends to Bob, Grangier said. The method works because amplitude and phase are non-commuting quantum variables, which means they, like the pairs of polarizations, cannot be measured simultaneously. "This can be used to hide the secret key," he said.

The researchers' scheme also contains a twist. Its process of comparing bits reverses Alice's and Bob's usual roles. "The trick of reverse reconciliation is that the correct key is not what was sent by Alice, but rather what was received by Bob," said Grangier.

Reversing the process incorporates transmission errors into the key, said Grangier. This is more secure because Eve can only intercept messages before they get to Bob. "Eve always knows less about what was received by Bob than about what was sent by Alice," he said. This gives Alice and Bob a larger advantage over Eve compared to standard quantum key distribution systems, he added.

The difficult part of putting together the scheme was evaluating how much information is shared by the sender, receiver and eavesdropper, said Gangier. "While this is simple and well-known for photon counting, it was not clearly done for [multi-photon systems] before our work," he said.

The researchers have proved the system is secure in some situations, but have work to do to prove that it is perfectly secure. The initial results look good, however, said Grangier. "Our feeling is that the protocol is also robust against more general attacks, but this remains to be proven."

The researchers are not the first to propose a multi-photon quantum key distribution system, but the other methods generally rely on unusual, hard-to-produce light beams.

The work is "undoubtedly... the most advanced [multi-photon] quantum crypto demonstration," said Nicolas Gisin, a professor of applied physics at the University of Geneva in Switzerland.

The distance the scheme can be used over is limited, however, said Gisin. "It is not clear how realistic the proposal is for significantly longer distances." More work also needs to be done before the security of the system is clear, he added.

The researchers are working on testing the protocol at the wavelengths used by today's telecommunications systems, said Grangier. "The first challenge [is] using telecom integrated modulators and optical fibers at [a wavelength of] 1,550 nanometers instead of our present bulk optics set-up at 780 nanometers," he said.

They are also looking into quantum repeaters, he said. "We're also interested in... quantum repeaters that would do for quantum key distribution what optical repeaters can do for usual telecommunications, i.e., reaching arbitrarily large distances [using] optical fibers," he said.

How close the scheme is to practical depends on exactly how it will be used, said Grangier. "Since quantum key distribution setups are now on the market, the problem is no [longer] only technological," said Gangier. "One should decide what is ultimately needed: high bit [rates] over short distances [or] smaller bits rates over larger distances; fiber systems or free space systems?"

Gangier's research colleagues were Frédéric Grosshans, Jérôme Wenger and Rosa Brouri from the Institute of Optics, and Gilles Van Assche and Nicolas J. Cerf from the Free University of Brussels in Belgium. They published the research in the January 16, 2003 issue of the journal Nature. The research was funded by the European Union Information Society Technologies program (IST).

Timeline:   Unknown
Funding:   Government
TRN Categories:  Quantum Computing and Communications; Cryptography and Security
Story Type:   News
Related Elements:  Technical paper, "Quantum Key Distribution Using Gaussian-Modulated Coherent States," Nature, January 16, 2003.


January 29/February 5, 2003

Page One

Data stored in live cells

Faster quantum crypto demoed

Bumpy surface stores data

Quantum computers go digital

Tiny hole guides atoms against tide


Research News Roundup
Research Watch blog

View from the High Ground Q&A
How It Works

RSS Feeds:
News  | Blog  | Books 

Ad links:
Buy an ad link


Ad links: Clear History

Buy an ad link

Home     Archive     Resources    Feeds     Offline Publications     Glossary
TRN Finder     Research Dir.    Events Dir.      Researchers     Bookshelf
   Contribute      Under Development     T-shirts etc.     Classifieds
Forum    Comments    Feedback     About TRN

© Copyright Technology Research News, LLC 2000-2006. All rights reserved.