How a Quantum Satellite Network Could Produce a Secure Internet

Last Friday, the Motherboard team of editors did a Reddit AMA. It was a good time. Lots of great questions were asked, one of them being to the effect of, What is the future of secure communication? Conveniently enough, I'd just finished asking one of the researchers at the leading edge of quantum communications research, Brendon Higgins, that very question.

The answer is simple enough: the quantum internet, e.g. an information-communication network utilizing quantum encryption in securing its transmissions. The question is pretty interesting, because we don't actually have secure digital communication now at all, and we never have. What we have is simply "good enough" security. The encryption keeping your credit card numbers, medical information, bank routing details, etc., relies only on current computing technology not being good enough to crack a certain type of very difficult math problem, the factorization of large primes. Weird, eh?

Higgins and his partner Thomas Jennewein published an article in Physics World a few weeks ago that proposes the next steps toward actual, for-reals secure communication. The new system, resting on a network of satellites to bounce encryption keys around, utilizes the principle of a quantum system that says it cannot be disturbed without both the sender and receiver of a quantum message knowing. Imagine intercepting a horseback messenger on some busted Middle Ages road--if you so much as glanced in his direction, let alone snagged his message for the King, he would drop totally dead.

That's about what quantum encryption is like; the slightest disturbance causes the system to self-destruct, alerting everyone involved. It would be the first totally secure communication system in human history. I emailed a bit with Higgins last week about the prospects for this future, what it might mean for privacy, and why we need satellites to do it.

Motherboard: I'm wondering if you can elaborate a bit on the need for quantum encryption. What will push it to where regular encryption schemes are insufficient? How are they insufficient now?

Brendon Higgins: It's an open secret that traditional encryption schemes like those that make secure internet communications easy fundamentally rely on the assumption that, basically, cracking them is a legitimately hard thing to do. The only justification for believing this is that, so far, we haven't found an easy way to do it, but there is no logical or foundational proof that this must be true. Worse, the exponential growth of computational power means that schemes that were arguably secure in the past can now be cracked using desktop computers, basically by sheer force, and this trend will continue into the future.

And then there are quantum computers looming on the horizon. Should such a device be built (and it's almost certain that, eventually, it will be), it could be used to run a new type of algorithm that would allow its user to crack several commonly-used encryption protocols with relatively little effort.

Quantum encryption techniques are different. They fundamentally rely on the inherent properties of quantum mechanics, one of the most thoroughly and successfully tested physical frameworks yet devised. Quantum encryption is not susceptible to the increases in computational power or advances in algorithms, and to crack a properly implemented quantum encryption device would be to crack a bedrock of modern physics.

It's easy to say things like 'a new quantum internet' and the like, but in the early days who will this exist for? What would the expansion process look like for quantum encryption becoming normal for regular old tasks?

Governments and banks--institutions that have investment in keeping certain information confidential--come to mind as likely to be the first consumers of quantum encryption networks. And we are already beginning to see this, with quantum cryptography having been employed to secure Swiss elections, as well as private expressions of interest in the development of the technology from numerous parties.

But a 'quantum internet' services more than just quantum cryptography. It would also allow for distributed quantum computation, i.e. the use of numerous and/or distributed quantum computers to perform information processing tasks with new algorithms that cannot be executed with regular computers. It's hard to say what applications this may ultimately lead to, as it's still early days, but we know that for certain interesting tasks--including e.g. integer factorization and database searching--this could provide significant speed-ups.

What are the biggest hurdles between us and a functional quantum satellite system?

It's interesting that the biggest hurdles for a quantum satellite system are not technological. We know that we have the capability to get such a system to work using existing technology. What doesn't already exist can be built; what hasn't already been demonstrated can be tested. (Of course there are some things that are more well-developed than others. For example, successful long-distance quantum cryptography has been demonstrated over a fixed optical link, whereas the necessary high-precision optical alignment system to operate over the varying ground-satellite link is an active development area.)

Rather than technology, the main hurdle at the moment is in raising leadership awareness and developing the motivation to put the resources towards actually constructing a device. One of our hopes is that, by demonstrating the feasibility of a real quantum satellite it will help to encourage new investment in the development of quantum networks and quantum satellites for commercial implementation.

I'm not sure I remember any cost estimates in your Physics World piece. How would it compare in the grand scheme of quantum computing research? Are we talking like the LHC of quantum computing here?

We see commercial interest driving development and deployment long before getting to the cost levels of something like the LHC. Certainly building and launching a satellite is not cheap, but it isn't massive-scale, and it is only one (albeit a significant one) part of the entirety of the quantum information field.

Something I've wondered for a while: with quantum encryption in the private sector, how will it relate to government eavesdropping, particularly backdoor-type requirements (like, say, there having to be a key available to the FBI or somesuch). Broadly, how are governments likely to respond to communication where a third-party key/eavesdropping is just impossible by definition?

I can't really speculate on what government agencies are likely to do in this situation, other than expecting them to try to do what they've always done. It might be difficult to mandate having backdoor keys available to any given TLA agency due to the way quantum encryption works (generating keys on the fly) but that cannot stop them from demanding backdoors in the electronics that drive such devices. Of course, this is ultimately true for any information technology.

Reach this writer at michaelb@motherboard.tv