E-mail updates
Follow on Twitter
Subscribe to RSS
Equinox Graphics
Clusters of entangled qubits, shown in this artistic visualization, could allow remote quantum computing to be performed on a server while keeping the contents and results hidden from the remote server.
If the future is heading toward "cloud computing," where most of your data lives on someone else's server, can you trust the cloud to keep a secret? Researchers say they've found a way to guarantee that your information will be secure in the cloud, using quantum entanglement.
The technique is called blind quantum computing, and it adds one more piece to a puzzle that could eventually be assembled into an entirely new infrastructure for data processing. Theoretically, quantum computers could outdo classical computers when it comes to making weather predictions, simulating biological processes, analyzing chemical reactions and, not incidentally, deciphering secret codes. Data security could become an even bigger issue than it is today.
Whom do you trust?
Today, most of your computing power probably resides on the device you're using, whether it's a desktop or a smartphone. If you send secure data someplace else, those bits are probably encrypted using classical mathematical techniques. They're tough codes to break, but they're not unbreakable. In fact, computer scientists say quantum computers might be well-suited for cracking today's classical codes.
At the same time, there's a trend toward developing devices that shift more of the computing power onto big servers. You would still use your tablet or smartphone or netbook for input and output, but the information is stored and processed as part of a huge cloud of bits on the server. That's the idea behind the much-debated cloud computing approach.
How sure can you be that the folks who manage the cloud won't meddle with your data? And could a malicious cloud client meddle with the central server? Such questions are tricky now, and they could get trickier if quantum computing takes hold, according to an international research team led by Stefanie Barz of the Vienna Center for Quantum Science and Technology at the University of Vienna and the Austria-based Institute for Quantum Optics and Quantum Information.
In this week's issue of the journal Science, Barz and her colleagues note that quantum computers will be so complex that there may be only a few of them in operation at specialized facilities around the world.
"A key challenge in using such central quantum computers is enabling a quantum computation on a remote server while keeping the client's data hidden from the server," they write.
Demonstrating blind computing
The researchers worked out a system to entangle photons of light that were generated by a nonlinear crystal, and then "process" those entangled photons on an experimental setup of beam splitters, filters and couplers. The photons served as quantum bits, or qubits, to be manipulated in two types of quantum calculations (Deutsch's algorithm and Grover's search).
In this scenario, the person who provided the qubits knows their initial entangled state, and can thus decipher the entangled outcome. But the company that does the data processing wouldn't know how the qubits were entangled — and thus could not even try to decode the qubits without essentially destroying them. As far as it's concerned, all those qubits look like a totally random hodgepodge. What's more, the system has a built-in verification scheme.
"By inspecting the output, you can know if the company really has a quantum computer, without disclosing your algorithm, the input, or indeed the output," the University of Oxford's Vlatko Vedral said in a Science commentary on the research. "The computation is thus 'doubly' blind."
Barz and her colleagues say there are still some technical challenges to be overcome. for example, it's theoretically possible for some of the photons emitted while preparing the qubits to reveal information about the "blind" phase of the calculation. Also, it's important to have a high-fidelity, low-signal-loss system for processing the qubits — whether they consist of photons with different polarizations, or electrons with different spins. But however the quantum computing puzzle is put together, the researchers say their experiments will have contributed a key piece.
"Our demonstration is crucial for unconditionally secure quantum cloud computing," they say, "and might become a key ingredient for real-life applications, especially when considering the challenges of making powerful quantum computers widely available."
More perspectives on the research:
- EurekAlert: Quantum physics enables secure cloud computing
- New Scientist: Quantum computer is blind to its own bits
- BBC: Quantum computing could head to 'the cloud'
- PopSci: Quantum computing will allow secure calculation
More about quantum computing:
- A quantum leap in computing
- Tales from the quantum frontier
- Spooky quantum entanglement disturbed
- Four-atom-wide wire may herald tiny computers
Alan Boyle is msnbc.com's science editor. Connect with the Cosmic Log community by "liking" the log's Facebook page, following @b0yle on Twitter or adding Cosmic Log's Google+ page to your circle. You can also check out "The Case for Pluto," my book about the controversial dwarf planet and the search for other worlds.
Leave a Comment:
Cloud compute at your own risk. Anything designed to secure data can be reverse engineered to do just the opposite.
I'd suggest reading a few quantum mechanics textbooks. After that, any material you can find regarding the integration of QM into modern computing and communication.
Perhaps then you'll understand where you went wrong with your post and why I marked it 'no value'.
Mitchell
This is the category of "electronic warfare". As someone put it a long time ago, "For every measure, there is a counter-measure."
Take this seriously!
Mitchell - I upvoted you (and agree with you). Me thinks if proamerica/john dunn read the article and do a little research about quantum mechanics then you'd have understood the entire point of the research: Any attack at the cloud sidewould require knowledge of the entanglement. Since there is no knowledge of the entanglement at the cloud side then one would have to employ an attack at the user end. If the attack is at the user end then the argument that your data was violated within the cloud is invalidated.
Of course this does require some rudimentary knowledge of quantum mechanics to accept that the qubit would be destroyed by inspection without knowledge of the entanglement, but at this point as much as our personal day to day lives seem to contradict quantum mechanics it is pretty well established to be the reality at the photon level. (If you don't believe in quantum mechanics, I'd sugest not using your GPS because the algorythms invovled in signal timing, etc. assume quantum mechanic effects to be real).
To expand upon what Mitchell is saying - if you didn't get it from the article, let me spell it out a little more clearly: It is physically IMPOSSIBLE to decode a quantum entangled encryption scheme (as far as we know at this point). The two entangled bits are moving around, seemingly at random - but they do this in the same direction at the same time no matter how far apart they are. At least that's how I'm understanding this with my limited knowledge on the subject matter, someone feel free to correct me if I slipped up on that.
It helps to read the article before posting.
As proamerica's post was completely in line with the article, I would say it is entirely unfair to mark it as having no value. I am no quantum physicist, but to assume that anything man-made is truly failsafe or foolproof puts way too much faith in man.
Now, if someone came in here saying "The only clouds I see are the ones between politicians' ears derp derp derp," I would safely consider that to have no value, and I would mark it as such.
lol, so you didn't read the article either? Again - physically impossible to "listen in" on a quantum entangled transmission. Just think of it as "magic", because honestly, most of the scientists still don't know WHY this works I'm sure.
It should be noted that the cloud could still mess up the calculation by trying to peek.
The random entanglement doesn't prevent something in the cloud server from trying to peek at the results, but if they do, they won't learn anything, and will produce a garbled computation that can easily be recognized as such.
Bottom line, you get the answer you sought or (if someone tries to peek along the way) clearly recognizable gibberish.
Seeing as peeking would possibly destroy the entanglement, that in itself could become quite a nuisance - hackers just going around and breaking that connection, then you have to wait to get a new system set up.
@breakingglass27:
Wait, you say you're not a quantum physicist just after you say that proamerica's post was completely in line with the article, after I specifically mentioned that not knowing QM was his problem in the first place.
I must admit, it takes a kind of bravado to throw yourself out there like that.
Mitchell
Hey now, all I was saying was that his post was relevant, whether or not it is correct. Simply correcting him (and me I guess, whatever) is all it should take, not condescendingly informing him his post had no value because he's not entirely familiar with the field.
Hehe, You know, you could've taken that last sentence in a good way.
Sorry, proamerica is sort of a one-post-wonder who wanders into the tech/sci forms and tends to just get a lot of things wrong. He actually deserves a lot worse than I gave him but since he's a 'one post then I'm out' sort of person he doesn't rate a more thorough post.
The brazenness of your post and how you said made me laugh which is why I chose the word 'bravado'. :-)
As for why his post is irrelevant, a couple of things. According to QM, entangled bits cannot be compromised. They are created together and stay that way with each other and regardless of distance, any change to one bit, the other also is changed instantaneously. This provides a secure form of communication as you cannot 'listen in on the communication' nor can you entangle either bit with an outside unit. The best someone could do is somehow untangle them, disrupting communication.
Secondly for the computer itself, this is a bit trickier as there are a couple different attributes at work. One is that observing a quantum state changes that state. Also you can have other bits rely on the state of each other. The above article touches on these and adds methods of encryption. Kind of what it boils down to is that unless you are using the exact method of interaction when it was set up, you cannot get anything other than garbage data if you try to access it. Why this is secure, is because we are relying on the physical properties of QM instead of your standard encryption on today's computers.
Proamerica is thinking that you can hack these computers, you can't, or perhaps listen on the communication, which you also cannot do. You can always steal the user unit, or something along those lines, but that has nothing to do with reverse engineering the tech.
Mitchell
Ah, I'm not familiar with proamerica's other posts, otherwise I might have caught the tone a little better. It's sort of like when, a year or two ago, I was confused why someone was being so outright rude to a user named Elemental1... of course it only took another post for me to completely understand.
And yeah, I'm still no quantum physicist, but appreciate the clarification.
I really do not see why anyone other than businesses would want to keep there personal data on the "cloud" where they will have to pay to store it and if you don't pay it would be deleted. The day computers can no longer store my data locally, will be the day I stop purchasing computers.
It could be a nice back up system. Keep your PC backed up through the cloud, so if say your house burns down - no problem, once you get your new PC, put in a code and it feeds all your data back down the pipe. Also, for people that travel a lot it'd be REALLY handy. All the processing power of a full sized desktop available through a netbook, laptop, tablet, etc, and you can access it wherever you travel.
Muskateer-802444 Just so you know it cost $$ for corporation to store data no matter where they store it.
And 640 KB of memory is definitely more than anyone would ever need, right? I see this claim being very foolish in the future.
Doesn't matter how secure. I will still want to keep my own stuff on my own computer. Cloud computing is just another "service" that eats into your budget.
Not sure that is true for all businesses. If you're a small business with no IT department but require a network with a couple of server it might be more cost effective to go cloud than to hire IT staff, hardware, and software.
I'm a bit confused how the entangled particles could actually be transferred, in the real world. The experiment used beam splitters and such, but a real life scenario would have to pass the data through the physical network, which these days involves a lot of routers, switches, wires, and fiber-optic cables, not to mention the cellular transmission portions. Wouldn't you need one continuous conduit (fiber-optic cable or whatever) in order to actually transfer the entangled particles to the user's computer? Yeah, I understand that the experiment demonstrates a concept, but will the reality ever emerge in practice?
Technically I'm thinking that each of the particles could be put into a storage device of some sort and transported to the desired location. The remain entangled regardless of distance - or they're supposed to anyway, the farthest I've heard of them doing this is 10 miles apart...
Yeah, but that was by directly beaming a light that distance, which required line-of-sight communication. At the Niels Bohr Institute, they worked on a method of entangling two clouds of cesium atoms, and "refreshed" their entanglement using lasers. Apparently, the hope is to do this kind of refreshing at intervals along a path that the cloud of atoms might travel, creating something like "quantum routers". I'm still skeptical at its practicality, at least in anything like the near term. But technology continues to improve and amaze, so who knows...?
So the entangled particles have to have a line of sight connection to eachother to stay entangled then? Or just for them to pass the traffic successfully? This quantum stuff is still a little over my head heheh :-)
No, line of sight is not required for entanglement, it was required because what they were entangling were photons that they wanted to beam a distance away. I suppose they could pass through mirrors or lenses, as long as what they pass through doesn't read/alter the entangled state (e.g., polarization, spin, or whatever). But entanglement tends to "erode" over time, especially as the number of entangled particles increases. Thus the need to "refresh" the entanglement in the NBI experiment, which entangled clouds of cesium atoms. (Not sure if I phrased that well, but that was the general gist I got of the NBI experiment.)
ahhh i see. Thanks for the clarification!
There are a lot of problems to work out. No one said this is going to happen any time soon.
Alan, just noticed a very similar article over on PopSci.com written by a Rebecca Boyle... relative of yours?
http://www.popsci.com/technology/article/2012-01/future-cloud-based-quantum-computing-will-stay-secure-using-secret-quantum-bits
How fitting this comes out on a day when the news is full of the post-SOPA blackout raid on megaupload and the method of filesharing. I would not be surprised to see clouds become the next method of illegally sharing pirated content by simply sharing access to someone's cloud among groups.
I don't see how we can avoid using clouds if we want to keep up in the pack. Perhaps a more progressive field of conscious commonality could use a road.
Who's to say some ridiculous federal law will won't be created forcing the cloud managers to divulge your personal data to the government ? Storage is cheap and I'll keep my personal data at home, thank you.
LOL I'm pretty sure you won't be owning a Quantum Computer anytime soon and this isn't going to be for the average home user, or average business for that matter. Also if your pc is on the internet it is not safe, you may think it is but think again.
It seems to me to be like this. First someone invents or improves cameras to digital and then the numbers of cameras in existence multiply a million fold and everywhere we go there is video evidence. Then many millions of us are on MSNBC and the soft ware writers begin to develop programs that predict what we like to read. Then you wonder when you see a article about someone killing and eating a cat in Arizona whether it and those other articles have been planted and maybe what articles are you not shown? Then you wonder if someone is keeping a record of what you are reading, who you send your emails to and if Directv is selling a list of what you watch to retailers. So, now, the commenters that are pro quantum mechanics tell us to trust them that they will not be making a people police or people control mechanism. Sounds like it is time to burn some books.