China on Tuesday launched the world’s first quantum satellite, which will help it establish “hack-proof” communications between space and the ground, state media said, the latest advance in an ambitious space programme.
…
“In its two-year mission, QUESS is designed to establish ‘hack-proof’ quantum communications by transmitting uncrackable keys from space to the ground,” it said.“Quantum communication boasts ultra-high security as a quantum photon can neither be separated nor duplicated,” it added. “It is hence impossible to wiretap, intercept or crack the information transmitted through it.”
As it happens one of our readers around here is a professor of quantum. More than a 0.2 professor of it as well. So as to the actual claim we’ll leave it to him to comment.
But my own suspicion is that the opposite will apply. That quantum will be easier to crack. Not, of course, because I know anything about the subject at all but because that’s the way I think the universe works. Yes, this is ludicrous physics but a spy plot would add in quantum entanglement. And Bond sneaks into the Chinese factory where they make quantums to put into satellites (and, obviously, qubits into chips and so on) and makes their supply of quantums ones entangled with ones at home in Bletchely Park. At which point we can read all their communications and calculations. And, as I think this works, they our so we only use this system to mislead them while communicating ourselves by tattooing the shaved heads of chavs. No one would notice anyway, would they?
Either that or we really do develop truly 100% secure comms and find ourselves interrupted by a bloke sitting on a cloud and shouting through his big beard “Get off the line!”.
Err, transmitting your keys from a satellite to the ground? I wouldn’t worry about whether your keys are guessable or brute forceable or not.
I’d worry about the receivable footprint.
Then all you* need to do is capture the keys and use them at random to decode intercepted Chinese comms. Once you get something that approximates to Chinese, you’ve got the right key and you can keep cracking that line until they rotate the keys.
* For a definition of “you” that is really Fort Meade and Cheltenham, with receivers based where-ever they need them under the satellite track.
BTW – we do have the ability to generate “100% secure comms”. One time pads.
The problem there is effort required for key generation (to make sure it has sufficient entropy) and key distribution (1 bit of key, to at least 2 locations, for every bit of data.)
QUESS is part of a collaboration between the Chinese Academy of Sciences and the Austrian Academy of Sciences: see http://www.iqoqi-vienna.at/home/research-groups/zeilinger-group/space-quantum-communication/ Definitely proof of principle stuff rather than actually real-life useful, but I don’t see why it wouldn’t work.
They seem to be distributing entangled pairs made in the satellite to two ground stations, which would permit them to use the Ekert 97 protocol or one of it’s relatives (much of their early work was BB84, which is not so exciting, but if they can teleport qubits they can run Ekert 97). The nice thing about Ekert 97 is that in principle it does not require that the entangled source be reliable: it can be error prone, or even compromised by an enemy. Tampering can be detected by the two ground users who can then stop communication. (In principle they could use entanglement distillation to continue communicating in the presence of low levels of tampering, but that would require rather more sophisticated technology.) As a general rule quantum cryptography can’t be tapped, it can only be blocked.
So I’m afraid Tim’s proposal wouldn’t succeed. That doesn’t mean that everything is secure of course: red-black leakage (accidental or induced by injection of trojans through a social engineering attack) and rubber hose cryptanalysis remain the obvious approaches.
Quantum crypto is one of those things – it has the potential to be uncrackable in theory, but the implementation side of it is not at all straightforward, and if you mess that up you think you’re uncrackable while leaking like a sieve.
@SE,
No keys are distributed by the satellite. The satellite distributes entanglement resources which the ground stations use to general one time pads for use as keys. The generation protocol ensures that the two ground stations come to an agreement on the same randomly generated one time pad, while nobody else (not even the satellite) knows anything about it beyond an upper bound on its length.
Wasn’t the most secure form of communications in WWII achieved by using speakers of an obscure language the enemy couldn’t possibly have knowledge of, in that particular case Navajo IIRC?
I once read about a Soviet dissident who learned about 30 languages to fluency. When asked why he learned Basque he said he was confident the Soviet censors would have no means of translating what he wrote or said in that language.
I reckon future communications only need a handful of Welsh, Geordies, and Glaswegians to achieve complete security.
“an obscure language the enemy couldn’t possibly have knowledge of”
Churchill sometimes used Hindustani during the War but although John Colville was amused by the references to “Lord President Sahib” he fretted that “I think the Germans know the language we use”.
Isn’t the problem with using quanta, that they are everywhere and yet nowhere at the same time. So the mere action of looking at the keys will change the message.
So the mighty Peoples Liberation Army arrayed along the Korean border gets the signal
“Spend three and fourpencen we’re going to a dance.”
Selecting th cryptographic method is about how long you need to keep something secret, the longer the more costly.
In the case of using Navajo, Welsh or Geordie that’s ideal for the battlefield when you only need about 30 minutes before your intentions are quite obvious. But I wouldn’t necessarily rely on it nowadays.
In the mid 80s in Celle, Germany, we got to know some German signals types quite well, ten pin bowling with them once a week. They were all fluent English speakers who had a strong grasp of idioms and no problem understanding our accents. It was almost like they deliberately got to know us for practice.
Tim N, some years ago I was acting as interpreter for an oilfield services company between a head office bloke (caricature Texan down to the cuban-heeled snakeskin boots, and with an accent to match), an Ulsterman (because the factory was in Belfast) and a Peruvian (where the product was being used). After about 30 minutes we gave up and went to the pub where, after a bottle and a half of Jack Daniels, we all understood each other perfectly.
Tim N: BiND has it exactly. Esoteric languages are great for real time communication in the field. In terms of cryptography, they are readily susceptible to pattern analysis. When you think about how codes are cracked (“AF is short of water”) it’s pretty trivial really.
Also, the Welsh for ambulance is Ambiwlans, which doesn’t seem impenetrable.
Ah yes, but the Welsh word for hospital is ysbyty. I’d love to see somebody hearing how it’s said for the first time try to write it down. 🙂
Quantum key distribution doesn’t use entanglement. https://en.wikipedia.org/wiki/Quantum_key_distribution
@J
That’s true of BB84, but the Ekert protocols do use entanglement. Take a look at https://www.amazon.co.uk/Quantum-Information-Computation-Communication-Jonathan/dp/1107014468