Quantum cryptography leaves the lab

[Aspman]

Article on the register thia morning giving some details of quantum crypto devices coming to the marketplace.

Quantum cryptography - long the stuff of cyberpunk novels and hi-tech spy stories - is leaving the laboratory and making its way into commercial markets. A briefing session at the UK's Department of Trade and Industry on Wednesday featured demonstrations of working quantum key exchange systems by QinetiQ, Toshiba Cambridge and US start-up MagiQ.

Quantum cryptography allows two users on an optical fibre network to exchange secret keys. It takes advantage of the particle-like nature of light. In quantum cryptography, each bit of the key is encoded upon a single light particle (or ‘photon’). Intercepting this data randomly changes the polarisation of the light, irreversibly altering the data. Because of this quantum mechanics effect any attempt by an eavesdropper to determine a key corrupts the same key. Quantum cryptography systems discard these corrupt keys and only use codes that are known to be secure. These quantum keys, once exchanged, can be used in a one-time pad.

Advances in quantum computers or the discovery of advanced mathematical algorithms might one day threaten conventional scrambling techniques but quantum cryptography, properly implemented, is immune from such attacks.

Professor Andrew Briggs, head of Quantum Information Processing Interdisciplinary Research Collaboration at Oxford University, said that the UK had played a critical role in research into quantum cryptography. He urged the British government to do everything it can to make sure British companies profit from this pioneering work. "We've made a world leading contribution to the underlying science. Britain should also be at the forefront of developing applications."

Professor Brian Collins, professor of information systems at Cranfield University, agreed. "The initiative is passing from scientist to system designers to exploit this technology. We ignore it at our peril; it may become the only show in town."

Consultancy QIP, supported by the DTI, organised two seminars this week designed to promote understanding of the potential impacts of quantum cryptography to government officials, reps from the financial services and telecoms industry and the media. It provided a rare opportunity to see a range of systems at work. The event itself was well organised and mercifully free of the marketing hype that often comes with the introduction of new technology.

Live and direct
Toshiba Research Europe demoed a prototype system that applied quantum cryptography to the transmission of streaming video, an initiative that might one day open the door to ultra-secure video conferencing. The system allowed each frame of IP video, to be encrypted with a unique digital key, creating an unbreakable sequence of IP traffic. Toshiba Quantum Key Server produces up to 100 quantum keys per second. Toshiba’s team has already achieved a number of world firsts including the a new type of light emitting diode (LED) that fires out photons one at a time and an ultra-low noise single photon detector.

US start-up MagiQ has already begun selling systems based on its technology in the UK. Staff from UK reseller NOW Wireless were on hand to show how their kit added an extra layer of security on VPNs. Once secure keys are exchanged, data can be encrypted using standard protocols. QintiQ demo-ed its free-space quantum cryptography system. Ultimately its approach might allow quantum keys to be exchange via satellite.

Swiss firm ID Quantique, along with MagiQ, are the only companies selling quantum key distribution systems commercially. Target markets are governments and financial services. Quantum key exchange systems from MagiQ sell at between $70,000-$100,000, a small premium on conventional cryptographic systems.

Although expense is not a factor several limitations with quantum cryptography remain. Hackers can't break codes protected by quantum crypto but they might be able to disrupt communications. Quantum cryptography is limited to use between two dedicated points, or perhaps around a star network. It can't be routed because this process would interfere with the exchange of keys.

Stuart Brooklehurst, SVP at Visa International, pointed out that the absolute secrecy offered by quantum cryptography only referred to the transmission of data. "It’s not a total solution. Risks are at least as great in other parts of the system," he said.

The big play for quantum crypto in financial services is for applications like data recovery and links between wholesale banks. Brooklehurst pointed out that systems have to go through expensive upgrades when protocols are upgraded. Quantum crypto, though initially more expensive, would provide the ultimate in investment protection. ®

Original story here:
http://www.theregister.co.uk/2005/04/28/quantum_crypto/

[sec_ware]

Hi

Just a follow up: Interestingly, a network protected by simple quantum encryption[1] has been broken, as one can read at nature news[2].
Nevertheless, it is clear to the authors:

Shapiro and Wong agree. And they add that a quantum cryptographic network can be simply tweaked to beat their attack. By making the key out of a lot of photons instead of just a few, the sender and receiver could ensure that the eavesdropper never got enough of the key to use it. Still, they say, the work shows that secrets — even quantum ones — are never entirely safe.

Important for some of us: commercial systems, like the one by idQuantique[3] are not vulnerable to this attack.


Cheers

[1] http://arxiv.org/abs/quant-ph/0702202
[2] http://www.nature.com/news/2007/0704...070423-10.html
[3] http://idquantique.com/

[Drakain Zeil]

So the real question is how many ways can we make imprints on a photon? xD
Brute force is probably not going to be added off-the-bat. You heard it here first.

[Aardpsymon]

Ok, the basic principle for quantum encryption relies on polarisation. A single photon vibrates in a single axis. Thus you can encode binary data on to photons by changing the axis of vibration. E.g. Vertical vibration is 1 and horizontal vibration is 0. You then detect the data by passing it through a vertical (or a horizontal) polaroid filter. If the photon is "1" it will pass through the filter and be recieved, if not it will be absorbed by the filter. Thus you build up a sequence of 1's and 0's.

Now the encryption bit. You change between encrypting on + axis and X axis at "random". Using | and / as 1, - and \ as 0. The receiving end knows your sequence of X and +, the key, and so can receive the data. Malice intercepting your data has to play a 50-50 guessing game with his polaroid filters. if he puts in a | filter and recieves a / photon there is a 50-50 chance that the photon will pass through the filter. Thus, every filter he gets wrong gives a 50% chance of receiving the wrong data. Additionally, when he sends the photon on he will encode the photon using the same axis as his filter, giving a 50-50 chance that Bob the recipient will receive the wrong data. In this way not only does Malice not intercept the correct message but Bob is very much aware of Malice's interference since the received message is corrupted.

Perhaps some day we will discover a way of identifying the vibration of a photon without destroying it, I don't deny that. The key point here is that the message is UNINTERCEPTABLE not UNDECRYPTABLE. Once you receive the message its no problem to decrypt it. In fact, its in the clear. But first you have to get the key (so, weakness right there....simply get the key).

Other points:
1) this requires a dedicated fibre line. If you have a dedicated fibre line between two points its difficult to evesdrop anyway. This is not something that gets sent over the net for anyone with a traffic sniffer in the right place to catch.

2) Its a bloody slow method of communicating. Sending single photons is slow especially allowing time for filter changes.

So yes, I do believe it is unbreakable encryption. I don't count getting the key as breaking the encryption. There is no way to "brute force" this. Its a laws of physics thing not a "10,000,000,000 years with the fastest PC ever" thing.