First quantum cryptography network unveiled

The first computer network in which communication is secured with quantum cryptography is up and running in Cambridge, Massachusetts.

Chip Elliott, leader of the quantum engineering team at BBN Technologies in Cambridge, sent the first packets of data across the Quantum Net (Qnet) on Thursday. The project is funded by the Pentagon's Defense Advanced Research Projects Agency.

Currently the network only consists of six servers, but they can be integrated with regular servers and clients on the internet. Qnet's creators say the implementation of more nodes in banks and credit card companies could make exchanging sensitive data over the internet more secure than it is with current cryptography systems.

The data in Qnet flows through ordinary fibre optic cables and stretches the 10 kilometres from BBN to Harvard University. It is encrypted using keys determined by the exchange of a series of single, polarized photons.

The first money transfer encrypted by quantum keys was performed between two Austrian financial institutions in April 2004. But Qnet is the first network consisting of more than two nodes to use quantum cryptography - a more complex challenge.

"Imagine making a phone call. If you just have one possible receiver, you wouldn't even need buttons," explains Elliott. "But with a network you need a system that will connect anyone on the network to anyone else." In Qnet, software-controlled optical switches made of lithium niobate crystals steer photons down the correct optical fibre.

Intruder detection

Quantum cryptography guarantees secure communications by harnessing the quantum quirks of photons sent between users. Any attempt to intercept the photons will disturb their quantum state and raise the alarm.

But Elliott points out that even quantum cryptography "does not give you 100 per cent security". Although quantum keys are theoretically impossible to intercept without detection, implementing them in the real world presents hackers with several potential ways to listen in unobserved.

One example is if a laser inadvertently produces more than one photon, which happens occasionally. An eavesdroppper could potentially siphon off the extra photons and decrypt the key, although no one has actually done this.

"However Qnet is more secure than current internet cryptography," says Elliott, which relies on "one way functions". These are mathematical operations that are very simple to compute in one direction, but require huge computing power to perform in reverse.

The problem is, according to Elliott, that no one has actually proved that they cannot be solved in reverse. "So who's to say that someone won't wake up tomorrow and think of a way to do it?"

 

Text 9

New robot face smiles and sneers

A new robot that, according to its creators, can express a full repertoire of human facial expressions was unveiled on Sunday.

K-bot has a feminine face and is capable of 28 facial movements, including smiling, sneering, furrowing her brow and arching her eyebrows. She also has cameras in her eyes to recognise and respond to humans.

David Hanson, of the University of Texas, Dallas, sculpted K-Bot's face using an electroactive polymer with 24 artificial muscles to provide facial movement. Her predecessor - Andy the android - had just four facial movements.

"This is the face for human robotics," Hanson told the American Association for the Advancement of Science's annual meeting in Denver. Previously, Hanson has designed robots for Disney theme parks.

 

Flexible skin

K-bot's face is made of a polymer developed by Hanson. It is a combination of an elastomer and a foaming agent, he says, which gives the appearance and flexibility of skin without requiring the complexity of human tissue.

The entire robot, which consists of the face, muscles and motors, weighs two kilogrammes. It cost $400 to make but Hanson says the cost would be even cheaper if the robot was mass produced.

He suggests K-bot could be a useful tool for scientists researching artificial intelligence: "The goal is to test adaptive intelligence systems which can recognize people and then respond."

Human problem

However, Cynthia Breazeal, at MIT's media laboratory, warns that the interactivity displayed by robots in films like Al and even C3PO in Stars Wars is a long way off.

Breazeal, who has researched how socially responsive robots interact with humans, says: "Science fiction is very difficult for us," because it gives people very high expectations.

"This isn't just an engineering problem," she says, it is a very human problem - we must also understand the social and psychological aspects of the interaction.

Text 10