Quantum crypto, from BBC

[Udhay Shankar N]

I haven't seen this discussed here yet. Is there something to this?

(Forwarded from Dave Farber's IP list, forwards elided)

>Dave
>
>Some IPer's may not yet have seen how close BBC says quantum cryptography
>may
>be.
>
>Bob
>
>  * Quantum leap for secret codes *
>British scientists are close to creating a system to keep data sent via the
>net
>secure, using quantum physics.
>Full story:
>http://news.bbc.co.uk/go/em/-/1/hi/technology/2963138.stm
>
>--
>
>Bob Rosenberg
>
>Quantum leap for secret codes
>Within three years, companies could be using quantum cryptography to protect
>sensitive messages.
>
>British researchers say they are close to producing an off-the-shelf system
>that exploits quantum physics to create a secure communications channel.
>
>The system encodes bits of information on individual particles of light.
>
>This week, the researchers demonstrated their system working over fibres 100
>kilometres in length.
>
>Key controls
>
>Secure ways of communicating are hugely important to companies as more
>business is done via the net.
>
>Contemporary encryption algorithms ensure sensitive data stays secret simply
>because it takes so long to work out which mathematical key was used to
>scramble the information.
>
>But companies could soon be turning to quantum physics to help them swap the
>keys they use to turn meaningful messages into scrambled characters.
>
>
>"It is a radically different approach to computer security," said Andrew
>Shields, leader of the Quantum Information Group at Toshiba's Cambridge lab
>where the encryption system is being developed.
>
>"Quantum cryptography allows users to swap a shared secret key that can be
>used to encrypt messages or authenticate messages sent across the network,"
>said Dr Shields.
>
>"We can absolutely guarantee that each key sent is secret," he said.
>
>Quantum cryptography works by using particular properties of individual
>particles of light, photons, to represent individual bits of information.
>
>"The sender sends the photons in a particular state which determines whether
>it is bit 0 or bit 1," said Dr Shields.
>
>The system being developed by the Toshiba researchers can form keys at a
>rate of up to 2 kilobits per second. Much higher bit rates will be possible
>in the future.
>
>Companies were unlikely to use the system to transmit entire messages, said
>Dr Shields, but it was perfect for sending the keys used to scramble and
>unscramble messages.
>
>Bits and PCs
>
>Dr Shields said photon physics meant eavesdroppers changed the properties of
>any photons they intercepted. As a result, it became obvious if a message
>had been read by anyone it was not meant for, he added.
>
>"The critical part is that bits have to be encoded on single photons because
>it is impossible to copy them. It gives them a way of testing whether
>someone else has read their message."
>
>Unfortunately, the delicate states of the photons used to represent bits
>also mean that quantum cryptography can only work along unbroken, and
>relatively short, fibre optic cables.
>
>By contrast, data sent across nations and oceans travels via fibre optic
>cables regularly broken by repeaters that boost signals to ensure they reach
>their destination.
>
>
>Dr Shields' team have demonstrated quantum cryptography working over
>distances of 100 km, which should be enough to cover large metropolitan
>areas such as London and Tokyo.
>
>Dr Shields said he expected the first users to be financial firms keen to
>establish secure communication between different offices.
>
>One weakness with the system was what happened to encryption keys once they
>started to be used in the leaky, insecure computers many of us had on our
>desks, said Dr Shields.
>
>"There are other vulnerabilities in the system that's true," said Dr
>Shields. "The key has to be used in the right way to ensure the security of
>the entire system."
>
>The Department of Trade and Industry has pledged cash to help the
>researchers refine their work and bring commercial quantum cryptography
>products to market.
>
>Dr Shields said commercial products should be ready in three years' time.
>
>The researchers will be making their work public at the Conference on Lasers
>and Electro Optics being held in Baltimore this week.


-- 
((Udhay Shankar N)) ((udhay @ pobox.com)) ((www.digeratus.com))


---------------------------------------------------------------------
The Cryptography Mailing List
Unsubscribe by sending "unsubscribe cryptography" to majordomo at metzdowd.com