Cambridge scientists discover way to build multi-user quantum key networks
- 05 September, 2013 12:51
The era in which everyday communications are protected from surveillance by the physics of quantum cryptography has moved a little closer with the news that scientists at Toshiba's Cambridge Research Lab have invented a way for multiple users to share one of the technology's most important but pricey components.
In a paper published in this week's Nature, the team headed by Dr Andrew Shields describe a Quantum Key Distribution (QKD) test network in which up to 64 users were able to share a single photon detector, removing the need for each user to have one of these expensive components each.
If it sounds like a mundane advance, as far as today's quantum cryptography (or QKD) technology is concerned it is anything but.
Conventional optical networks are highly efficient because they work on the principle of a shared infrastructure that can handle numerous data streams and many users at once without complication. To date, QKD systems have had to set up a single point-to-point link for every connection, requiring duplicate photon detectors for every user wanting to receive a message.
In making it possible for multiple users to share a single detector, the team's innovation has been to discover a way that each user's stream of photons can be distinguished from one another, a complex feat for QKD. The scientists also had to do this while taking account of incredibly small thermal stresses on the optical fibres that can generate errors.
Although the advance gives hope that QKD systems will in the near future start looking a lot more like conventional communications networks, there are still a few limitations. Data rates at the 64-user level tested so far are very low by conventional standards; getting bitrates to a modest 250kbps over a 12-hour test run meant lowering the number of users to only eight.
On the other hand, "continuous operation over a month would allow unconditionally secure one-time pad encryption of more than 10 GByte of data for each user, which is enough for example to protect over one hundred thousand emails," the research paper said.
Since they burst out of the lab a decade ago, QKD systems have only slowly made it into commercial operation with the military and government, or basically anyone that is for whom security is an absolute must. The technology remains mired in high cost, high complexity and low bitrates which tail off over distances.
But developing the equipment so that QKD systems start to resemble the topologies used by established technologies is starting to advance.
"There is work to do making it compact and reliable," admits Dr Andrew Shields of Toshiba's Research Labs. "This will take time to do but it's achievable." The short-term challenge for the technology was to make it as cheap and practical as possible, he added.
The latest development was necessary to allow QKD to scale to large numbers of users, giving it a means of getting beyond today's niche applications, he said.
Advances have been incremental, such as a previous breakthrough that allowed QKD to share an optical fibre with conventional wave division multiplexing optical communications.
QKD's lure remains very strong because it offers a way to harness the principles of quantum physics to send data between users such that any interception or surveillance cannot be hidden as it can today with photo-electronic technologies. The NSA doubtless has a great interest in such systems - as long as nobody else is using them.