With AU$3.26m from Australia's government, quantum crypto outfit QuintessenceLabs has set to work getting the fibre out of its diet, and instead running quantum key exchange over free space.
While there have been scientific demonstrations of quantum exchanges both terrestrially and between earth and satellites, QuintessenceLabs' project is towards product development – its objective is to deliver up technologies and proof-of-concept data, to make the technology ready for commercialisation.
The company's CTO John Leiseboer explained to The Register: “The intent of the project is to take the technology that we have been working with – and take some of the new research we're applying to improve capability – and bring that to a point where it is ready to be productised and commercialised,”
“The outcome won't be a product, but a technology that is ready for productisation.”
Quantum key distribution (QKD) uses quantum states to swap encryption keys between two parties. This provides a tamper-proof, eavesdropper-proof channel for the key exchange: any third party (say, a man-in-the-middle) degrades the quantum channel and the key exchange therefore fails.
Once the two ends of the transaction have swapped keys, they use traditional encryption on conventional communication networks (which have much higher bandwidth than the QKD system).
The grant to develop free-space QKD comes from the Department of Defence's Defence Innovation Hub, and once complete, the resulting systems will be deployed in Defence and other sensitive government networks.
Why free-space QKD?
The company is one of several already offering (QKD) systems that operate over fibre networks, so Vulture South asked Leiseboer what's the value in a free-space solution.
“There are two main drivers that we're looking to put free-space use into. One is to give us to global coverage – having free-space QKD up to a satellite network extends beyond the reach of terrestrial links.
The other is in ad-hoc networks, he said: “Think of a forward-deployed airbase, and troops within view of that, not having to run fibre / wire to deploy QKD.”
In a state
Most explanations of QKD (including this author's) focus on using individual photons to exchange information: create a photon under the right conditions, and it splits into two entangled photons. That entanglement is the key distribution channel between Alice and Bob; and if Eve tries to eavesdrop, she spoils the entanglement.
Single photons are fragile, though: the discard rate is so high the successful Japanese experiment got a one-in-ten detection rate.
QuintessenceLabs takes a different tack, instead modulating quantum information onto the amplitude and phase quadratures of a laser, because that's a much easier channel to track and detect than discrete photons.
Leiseboer explains that unlike using phase and amplitude to carry classical information, the variations creating quantum modulation are very small.
“If we can modulate at an extremely low level both the phase and the amplitude, we've placed on the carrier a quantum-modulated signal”, he told us.
That makes noise the enemy, because “the thermal noise in the components can swamp the quantum signal.”
In this scheme, what's achieved is referred to as a “virtual entanglement” – the proof of security is based on “true entanglement”, and the equivalence of QuintessenceLabs' operation to entanglement relies on a mathematical proof.
“We create the quantum state at Alice. She knows what that state is, and sends a quantum signal to Bob.”
Since Alice and Bob know the quantum states (phase and amplitude) and the state of the channel (for example, its noise), they can then use Heisenberg's uncertainty theorem to prove the channel is secure, because “you can't measure both of those related variables (phase and amplitude) to an arbitrary accuracy.”
Any departure from what Alice and Bob know they can measure implies the presence of Eve trying to snoop on the channel, and the measurements are discarded.
And what Eve gets is useless anyhow: “Only Alice and Bob can get the information; Eve will always have an error that means the information is of no value to her”.
This project, Leiseboer said, will focus on three challenges: cost, performance, and reliability, because those are the “three areas that will take the most effort in commercialisation”.
“The program is a three-year program – during those three years we have a number of milestones that have to be delivered, roughly along the lines of refining the theory, design; doing the development of proof-of-concept technological components; then doing the experiments and delivering the results.” ®