UK crypto boffins turbo-charge quantum cryptography
Photodiodes get go-faster stripes
Boffins at Toshiba’s Cambridge Research Labs have developed a technique for turbo-charging the speed of quantum cryptography transmissions.
Quantum key distribution (QKD) systems currently operate at dial-up rates of less than 10 kbps for 20km, making them suitable only for point-to-point links. The Cambridge researchers have developed a technique for boosting these speeds by a factor of a hundred, increasing rates of transmission to above 1Mbps.
Crypto researchers have been able to boost transmission rates from the dial-up to broadband range. They used a single photon detector developed by Toshiba, which can work 100-times faster than earlier devices. Toshiba explains:
Current QKD systems are limited by the semiconductor devices (avalanche photodiodes) used to detect the single photons. One photon triggers an avalanche of millions of electrons in this semiconductor device which can be sensed by electrical circuitry in the QKD system. The problem in present systems is that some of these avalanche electrons can be trapped in the device and later stimulate a second spurious detection count. As these noise counts cause errors in the key, current detectors must be operated with long dead times to allow the decay of any trapped electrons. This limits the clock rate of current QKD systems to around 10 MHz and thus the secure key bit rate to under 10kbps for a 20 km fibre.
The Toshiba team has devised a method to detect much weaker electron avalanches. This strongly reduces the chance for an electron to be trapped, allowing the detector to be operated at much faster rates. Using this method Toshiba has increased the clock rate of their QKD system to 1.036 GHz, approaching the value used in conventional optical communications. This allows a raw bit rate of 9Mbps over 20 km of fibre, which in turn provides 1.02 Mbps of secure key. This is the first time that a secure bit rate exceeding 1Mbps has been achieved.
Although Toshiba's advance offers the possibility of creating kickass quantum crypto networks that support multiple users, the technology is unlikely to be commercially available for some time. Toshiba researchers have only demonstrated the advance in the lab and are yet to build even a fully functional prototype.
Nonetheless Toshiba said its technique allows a form of encryption that is theoretically unhackable for optical fibre networks connecting far more users than was previously possible. More information about Toshiba's research in quantum crypto can be found here.
News of Toshiba's research comes alongside the launch of a quantum cryptography network in Vienna, Austria, as part of a collaboration between crypto boffins across Europe on the SECOQC project. This network is based in part on older, lower-speed technology from Toshiba, which makes it unable to support more than eight users on a quantum cryptography network at any time.
Quantum cryptography systems are designed to allow the exchange of secret keys which can then be used in one-time pads, creating a theoretically unhackable system for the secure exchange and authentication of messages. The polarisation of individual light photons is used to signify each bit of a key.
The rules of quantum mechanics mean that any eavesdropper who intercepts this stream of photons will be unable to simply relay the data stream intact. Because of this effect, any attempt to eavesdrop on a quantum cryptography key would come across as a corrupt transmission across a unacceptable noisy link. Such keys would be dropped and transmission repeated until exchanges without interception take place. ®
Although this is rather impressive (despite the research behind it having been available for a long time), it's not really "quantum cryptography". It's using quantum physics to distribute a value. The cryptography used is still the same rsa/dsa/shared key/whatever. Thus surely it's still susceptible to the same attacks - if quantum computers "make it", then brute-force attacks on ciphertexts become simple.
Of course, the problem is lessened by using the keys as one-time pads, but still, cracking messages is possible, surely.
Point to Point?
Er, I'm not an expert in this but I thought quantum crypto was necessarily point to point? By reading the photons you drop them into a known state, which means they can only be read once. It's why quantum crypto can't be intercepted without detection.
Yes, it's incredible secure, but only so long as your network doesn't require any switches or routers, which kind of limits its application. 10kps or 1000kpbs, if you're limited to 25km of point to point it's really not going to make a great deal of difference.
"If they manage to increase the bitrate further one will soon be able to transmit VCD quality pr0n videos in complete security. Who knows, maybe they will even manage to achieve DVD-quality soon!"
With the new obscene publication laws we'll probably all be using quantum cryptography by next year! Just the boost people needed to take this all underground!