What do you get when you combine space, lasers, photons, the laws of physics, a Fortune 100 company, the Canadian Space Agency and a scrappy space startup?
The answer, it is hoped, will be a revolution in encrypted communications. Or, at least, the start of one: a mission to test quantum security in space. Why might you want to do that? Let me explain, with the help of a scrappy space startup and a seriously clued-up quantum security boffin.
Meet the QKD in space players
The Fortune 100 company involved here is Honeywell, the prime contractor for the Canadian Space Agency’s Quantum Encryption and Science Satellite mission, QEYSSat. The aim? Quite simply to put space-based quantum key distribution (QKD) to the test. More of that in a moment, but first, let’s meet the scrappy space startup.
Loft Orbital is a company that specializes in deploying and operating space infrastructure as a service. Using its Payload Hub technology, Loft Orbital takes a “Yet Another Mission” or YAM approach to payloads with a hardware and software stack to enable plug and play sensors on a standard microsatellite platform.
QEYSSat is, I am informed, the largest contract since Loft Orbital was founded in 2017. By coincidence, the same year that the Chinese Academy of Sciences launched a similar QKD program using the Micius satellite.
Quantum key distribution boldly goes where the Chinese have been before
So, why should you give a rat’s behind if it’s all been done before? Because, dear reader, QKD is a nascent technology, so every new test program will, almost inevitably, unlock further and valuable information. A few years is a very long time in quantum technology, to bastardize the political idiom.
There are a bunch of differences between the older Micius approach to QKD and that which QEYSSat is taking. For a start, QEYSSat is aiming to be less than 20% the size of the Micius satellite and will leverage commercial technology. Hence the involvement of Loft Orbital. Does size matter? You betcha. Reductions in size of that scale should lead to significant savings in both cost and time as far as the next generation of test projects is concerned. Size and mass will also be key if you’ll forgive the pun, as any QKD implementation at scale will demand a large satellite constellation.
Ultimately, if all goes according to plan, QEYSSat could have broad-reaching impacts as it should prove the capability to deliver QKD over much longer distances than the current ground to ground tests have managed to date. “This mission will demonstrate game-changing technology with far-reaching implications for how information will be shared and distributed in the future,” says Loft Orbital CEO, Pierre-Damien Vaujour, “we are honored and thrilled to be supporting it.”
Time, I think, to bring in my friendly quantum security expert, mathematician and security researcher, Dr. Mark Carney, who you may remember helped me explain why the math says Person Woman Man Camera TV made such a lousy password. Dr. Carney has a particular interest in quantum key distribution threat modeling, so makes the ideal guide to what we can expect, or not, from the QEYSSat mission.
How does quantum affect security?
“There are four ways quantum affects security,” Dr. Carney begins, “quantum computers break classical algorithms, post-quantum algorithms try to get around this by using harder math problems in classical crypto, quantum algorithms can be used to accelerate decisions (popular in quantum finance, but nobody in infosec has really looked at what algorithms can help where), and QKD, that uses quantum effects to do cryptography, bypassing the need for ‘mathematical crosswords’ altogether.”
Still with me, good? Because it gets a little more complicated from this point on.
The algorithms that drive QKD are oldish, and the most popular and well-established, BB84 and E91, primarily work in the same way.
- Alice sends photons with random polarizations
- Bob measures them with randomized polar filters
- Some exchange takes place to correct errors that occur naturally in the physics
- Alice and Bob use the output key in classical cryptography, knowing that they weren’t vulnerable to any issues with, say, ECDHE key exchange.
“Because regular cryptography goes over regular networks, it is fully error corrected,” Dr. Carney says, “the security is in the underlying math. As such, it can be packet-switched without any consequence.”
QKD in space brings what to the security party, exactly?
What has all this got to do with QKD in space? I’m getting there, and so is Dr. Carney. “The problem with QKD is that packet switching is somewhere between very very hard and basically impossible,” he says, “because unlike the security of classical crypto being in the math, the security of QKD is in the physical photon state.”
Time to get your ‘just accept this at face value’ head screwed on: if you observe a photon, the quantum effects you are using disappear and you may as well just use classical crypto because it is much better at being transmitted in the clear.
So, if not packet switching, then what? “You need a direct fiber link to do light photon-based QKD between every single endpoint you want to exchange a key with,” Dr. Carney explains. One major manufacturer of QKD fiber solutions produces building-to-building link equipment so that the internal security of the network is the only concern of the QKD keys produced. “This is where satellites turn out to be really handy,” says Dr. Carney, “send up one satellite, and have a load of users communicate with that, and no need to build dozens or hundreds of fiber links.”
If you have a laser array and a laser receiver, you can send pulses of photons up to satellites and still do QKD, albeit with higher error rates due to atmospheric diffusion of light that cannot be avoided. Dr. Carney will come back to that shortly, I’m sure.
“Another advantage of space is that you don’t need fiber repeaters,” he says, “and for distances of over 14km, single fiber connections get kind of useless.” There are fiber repeater network designs for QKD, but these are not necessarily immune to tampering, so breaking the trust modeling according to Dr. Carney.
The downsides of taking quantum into space
“I mentioned error and atmospheric dispersion on uplink before,” Dr. Carney reminds me, “as much as bad weather doesn’t actually affect cloud computing, cloud cover certainly affects QKD! Dispersion on the way down is also an issue, and targeting your downlink comms is also hard.”
It turns out that getting the aperture of that link down to a minimum seems like a tough problem. “I don’t think the calculations are favorable if your downlink laser disperses over a broad area,” Dr. Carney adds, “Eve would just have to plant a small mirror on your fence or carefully park another satellite quietly next to yours,” to break the threat model once more.
Dr. Carney is of the opinion that “going into space solves a few problems, but also introduces others.” Not least because QKD has a fundamental problem which is hard to solve under any circumstance: all of the security is in the physicality of the system. “One foot wrong,” Dr. Carney says, “and you can fail pretty badly very quickly.”
Where does all this leave the future of QKD?
As for the Chinese Micius program and what that taught us about QKD in space, the latest I heard was a June 2020 paper published in Nature that explained “entanglement-based QKD between two ground stations separated by 1,120 kilometers at a finite secret-key rate of 0.12 bits per second, without the need for trusted relays.” That paper claims the methods used increased the on the ground secure distance tenfold and increased the “practical security of QKD to an unprecedented level.”
And what of Loft Orbital, which seems to think that this new QKD technology should be available to the private sector, and adopted at scale, in the 2030s? Dr. Carney doesn’t have a problem with that as a date for adoption, given that Loft Orbital is demonstrating how microsats are getting ever easier to launch.
“Adopted at scale,” he says, “this is I think the kicker. There seem to be a lot of variables in the mix that don’t have easy engineering solutions. Unless you are launching a satellite per region and getting decent coverage with superb bandwidth to mitigate issues such as cloud cover, it’s hard to see how the cost viability is maintained.”
One thing is for sure, this is a move forward, and it will be interesting to see where all this takes us. Especially with “private equity making investments that heretofore were only really of interest and in reach of nation-states,” Dr. Carney concludes.