Could networks be made totally secure? A technique based on the optics of single photons, called quantum cryptography, promises to do exactly that, by allowing users to tell if their data has been tampered with.

What's more, a recent experiment conducted by scientists from the University of Geneva, and its spinoff company id Quantique suggests that quantum cryptography is becoming robust enough to use in the real world.

Damien Stucki and his colleagues reported the distribution of a quantum "key" over a 67 kilometer standard fiber link between Geneva and Lausanne in Switzerland. The key is the information needed to code and decode a secure transmission. They reported the work in the New Journal of Physics on July 12.

Before digging into the details of the experiment, let's look at the basics of quantum cryptography, or, to give it a more accurate name, quantum key distribution. Imagine that two people -- usually called Alice and Bob -- want to have a secure conversation. First they must transmit a secret key for encoding and decoding their transmissions. The problem is how to tell if this information has also fallen into the hands of an eavesdropper, Eve.

Quantum cryptography gets around this problem by using single photons for transmitting the key. Since a single photon cannot be created or destroyed, any eavesdropping or tampering is immediately apparent. If Eve attempts to intercept or copy the secret key, it introduces errors into the transmission that Bob receives, so he learns that the key is unsafe.

Right now, the most common method of encrypting transmissions on the Internet is the so-called public key. Alice sends Bob a public key, which allows him to encrypt information before sending it to her, over a public channel. She retains the related "private" key, which allows her to decrypt the information when it arrives. Security depends on so-called "one way" mathematical functions, which are easy to compute but very difficult to de-compute.

Unfortunately, there is no proof that these kinds of code are secure -- sometimes it is often simply a matter of time before these codes can be cracked, as computer processors get ever faster. If scientists ever manage to build a quantum computer, which can perform many calculations in parallel, then the available computing power will totally destroy the security of public key systems -- and that's a not insignificant worry for security application developers.

But although quantum cryptography could be the answer to current and future security problems, it's difficult to implement in practice. For a start, it uses single photons, so it requires light sources capable of generating single photons, and ultra-sensitive detectors capable of recognizing single photons. Noise is a big headache, so detectors have often been supercooled -- which requires expensive, bulky equipment.

The Swiss researchers say theirs is the first experiment to use "plug-and-play" equipment, without nitrogen coolers or continuous operational adjustments. The transmitter generates bursts of optical pulses at 1550 nanometers. 0s and 1s are represented by different "phases" of the pulse, and are created using an interferometer -- a standard arrangement based on simple optics like a beam splitter and delay line.

Both transmitter and detector are housed in 19-inch rack units, and each is controlled by a PC. In the commercial product, which id Quantique hopes to release later this year, the computers will be integrated into the boxes.

The system should work over any commercial fiber network, according to the researchers. To prove it, they have tested their system on a variety of fiber optic links around Geneva and Lausanne, the longest of which was 67km underneath Lake Geneva.

However, the system has a distance limitation of around 70 km, because of the attenuation in optical fiber. It's not possible to send the key through Erbium Doped-Fiber Amplifiers (EDFAs), which would alter the physical properties of the optical signal.

For the full paper, click here: http://www.iop.org/EJ/S/UNREG/abstract/1367-2630/4/1/341/

— Pauline Rigby, Senior Editor, Light Reading
http://www.lightreading.com