digg

Facebook Connect Join Digg About

Quantum Encryption

businessweek.com — Scientists have created an unbreakable cypher through the use of quantum physics. "Uncertainty is the principle we exploit," says MagiQ's VP for engineering. "It's impossible to find the key, because the photon can be measured once and only once. An eavesdropper can't measure it, and so can't get the key." Props to Heisenberg!

Who dugg this? Made popular Nov 9, 2006

Digg DudeWhat is Digg?

Digg is a place for people to discover and share content from anywhere on the web. Digg surfaces the best stuff as submitted and voted on by the community. Take the Tour to learn more.

68 Comments

show profanity settings
expand all
  • RckmRobot, on 10/12/2007, -2/+35This is not quantum encryption. It's quantum key exchange, and it is by no means a new idea. The encryption method simply makes use of a one-time pad, and the quantum part of it is only used in delivering the pad securely, not for actually encrypting anything.

    The biggest downfall in this is that it still requires classical (read: non-quantum) communication to make use of/create the one-time pad. This leads to more of a possibility of non-security.

    Regardless, it's still a cool use of technology and quantum physics.
  • DeusMachinae, on 10/12/2007, -4/+28You'd need to have GF for that, so your porn is safe.
  • mikesherov, on 10/12/2007, -2/+9Or they might know everything about the speed and nothing about the location.
  • cathars1s, on 10/12/2007, -0/+6It reminds me of the time Heisenberg was caught speeding. The policeman asked him, "sir, do you have any idea how fast you were going?". Heisenberg replies, "No, but I know EXACTLY where I am!"
  • DeusMachinae, on 10/12/2007, -2/+8I don't think BusinessWeek.com is going down anytime soon.
  • inactive, on 10/12/2007, -3/+8TrueCrypt. If your gf demands the pw, give her the fake one that has personal letters to yourself about how much you like her.

    That will score you major points.
  • cmiz, on 10/12/2007, -1/+6@people that think this won't be cracked:
    in theory, practice and theory are the same... in practice they are not. this system is perfect in theory, but perfect in practice is a rarity if not an impossibility.
  • thorseth, on 10/12/2007, -0/+5If someone else has been listening in on the quantum channel the key will be destroyed. The receiver will know this and can try again. The classical communication is worthless without the quantum information, this is not the problem. The transportation of undisturbed quantum states is the problem. The fiber connection in the article has a range of 140 km yes, but the transmission quality falls exponentially with distance - and this is still true for linked relays.

    I don't think we will see a q-net right away, but one can always hope.

    jjesusfreak01 beat me
  • jjesusfreak01, on 10/12/2007, -2/+7You dont get it, it cant be cracked. There are two sides to this. The first is the encryption key, which is as long as the message, making it uncrackable (there is nothing to compare, every single bit sent is encrypted separately) without the code. What quantum cryptography does is simply create a method by which the keys can be exchanged securely. By using this method, you can know if anyone tries to intercept the key stream, and if they do, you just find thr source of the leak, and retransmit a new key, ditching the old one. The point is that either you or the person intercepting will end up with the key, and you will know if you dont have the whole thing.
  • inactive, on 10/12/2007, -2/+6That's why I come to Digg, to find articles about 20 year old technology.
  • Homunculiheaded, on 10/12/2007, -4/+8Nah, whoever came up with that rule doesn’t even know a thing about the speed and location of an electron.
  • rnelsonee, on 10/12/2007, -0/+3Honestly, there's no *conceptual* way this encryption can ever be broken. I understand why you made your comment - because every scheme (aside from One Time Pads) can be broken, and they have been (just about - or can be with enough computing power). But quantum is completely different. Either RTFA or read better ones. Personally, I've seen much better articles out there. The (very) basic idea is that it is a OTP, but the pad is part of the message. In traditional means, this is a disaster, as an eavesdropper can just look at the OTP. But in quantum mechanics, the data changes depending on how it's viewed. Unless "Eve" is using the same equipment/computer that "Bob" is using, she cannot determine the pad used.

    I wish I could find the article I read years ago that had a great picture of the setup, but these articles seem to hit on it pretty well:

    http://searchsecurity.techtarget.com/sDefinition/0,,sid14_gci284012,00.html
    http://library.thinkquest.org/C008537/quantum/cryptography/cryptography.html
  • thorseth, on 10/12/2007, -0/+3A working quantum computer would break even the longest classic cypher in a very shot time. This is why the NSA and other spooks are very interested in this technology and in quantum computing. How long are the secrets that has already been transmitted safe, 20 years or 50 years?

    And by the way it is _not_ the same as classical encryption.
  • NoSalt, on 10/12/2007, -2/+5I can break it ... all you need is to use a "Heisenberg Compensator" and it'll crack.

    ;-)
  • zeeeej, on 10/12/2007, -1/+4I was getting worried. We changed the time required to break a strong key from a billion years to "forever." Whew!
  • thorseth, on 10/12/2007, -0/+2These guys can send the key 140 km - longest I have heard of.
  • ostracize, on 10/12/2007, -0/+2How do you like them apples Eve!
    - Bob & Alice
  • Lobster, on 10/12/2007, -0/+2Our motto is "It's Impossible - let's do it"
    In order to eavesdrop on an entangled quantum transmission, you alter it. Thereby alerting the compromised system. Our solution is to tap a line in a parallel dimension, where the data will be compromised - but who cares . . . we get the data and no one knows. Concept is there, application is VERY advanced.

    Multidimensional hacking will be stopped by the Chronology Protection Agency.
    http://tmxxine.com/Wikka/wikka.php?wakka=ChronologyProtectionAgency
  • SenatorPenguin, on 10/12/2007, -0/+2Actually, the encryption isn't compromised by the standard transmission of the actual message. One-time pads, if done correctly(completely random bits the length of the message used only once), are impossible to crack, as every other message of the same length can produce the same encrypted message. As far as I can tell, the biggest vulnerability is the exploitation of quantum physics, which doesn't follow normal rules, and any new developments in that area could change how QKE works.
  • OneZeroZeroOne, on 10/12/2007, -1/+3No, this isn't crackable, people.

    However, it is VERY limited in it's practicality. All it allows you to do is determine that someone is eavesdropping on the transmission.

    What's worse: having a conversation and not knowing if someone has listened in, or not being able to have a conversation at all?

    Using this technique, you know that someone is listening and therefore you have to keep "handshaking" with your intended target until the eavesdropper goes away...which may be never.
  • Spetz, on 10/12/2007, -0/+2A quantum computer would have the same difficulty cracking this as normal computers have cracking normal ciphers.

    Quantum computing, if it can be made applicable, should render all current ciphers obsolete.

    And yes, there was an article a while back about this in New Scientist.
  • rnelsonee, on 10/12/2007, -0/+2No, you don't have to keep handshaking - that's the whole point. It's acceptable to have someone eavesdrop on they key exchange, because the intended recipient is changing their polarization filters in a (pseudo)random way, and unless the eavesdropper changes his filters in the exact same way as the recipients, he will not be able to determine the correct key. Hell, the sender doesn't even get the full key because he doesn't need it.

    Unfortunately, this article stinks. But other ones (not found in Business Week) explain this in more detail.
  • rnelsonee, on 10/12/2007, -0/+2FWIW, it *does* use the Uncertainty Principle...
  • DeusMachinae, on 10/12/2007, -0/+2sudo randum numbers, is that a UNIX command?
  • MaSC, on 10/12/2007, -1/+2If you think the Heisenberg Uncertainty Principle is scientific babble then you really have no idea what you are talking about. The basics of the Uncertainty Principle are very simple.
  • cathars1s, on 10/12/2007, -0/+1Or as Aleksey Vayner would say, Impossible is Nothing.
  • thorseth, on 10/12/2007, -0/+1The two statements are equivalent. The measurement of one quantity "disturbs" its complementary quantity (momentum & position) so observation and uncertainty is linked together.
  • vuke69, on 10/12/2007, -4/+5RckmRobot:You are 100% correct.

    Now if they could only exploit a quantum connection, and/or quantum tunneling to do the actual transfer. That would be the ultimate. Unfortunately it will probably also be impossible for the foreseeable future, barring an incredible breakthrough in quantum physics (not to mention you would need a gigantic loophole in the uncertainty principle for tunneling to work AFAIK).
  • cathars1s, on 10/12/2007, -0/+1The only type of encryption worth having is that which can be open-source and still indecipherable. Things like the Enigma code are an artifact of the past.
  • Spetz, on 10/12/2007, -0/+1I'm talking super computers. You can number crunch most internet encryption, but it just takes forever and you cant do it unless you're the NSA. So most users have nothing to worry about.
  • anonydigg, on 10/12/2007, -0/+1"quantum key infrastructures are vulnerable to double man-in-the-middle attacks"
    I think what he means is,
    read it, duplicated it, send it. The newly transmitted signal has not been intercepted anywhere, and is identical to the original signal.

    btw I think they misrepresented Heisenberg's principal. The idea presented is correct, but it's not what Heisenberg said. He never said anything about measurement errors/alteration as a result of it. It simply refers to inherent properties of sub-atom particles; read Wikipedia on this misconception.
  • NickSumner, on 10/12/2007, -0/+1Just as existing public key infrastructures are vulnerable to man-in-the-middle attacks, quantum key infrastructures are vulnerable to double man-in-the-middle attacks. The probability of this can be very low, but it is inherently part of all existing quantum key designs.

    It may not be crackable, but with man-in-the-middle, you don't need to crack it; you just need to deceive.
  • evgen, on 10/12/2007, -0/+1No. It is only useful for a single point-to-point link. Once you hit the first switch/router/hub then the data is no longer protected by quantum encryption. This is one reason, among many, that quantum encryption will remain in the quasi-snake-oil realm that it has always occupied.

    It is cute, but utterly useless for 99.999999% of the real-world applications of cryptography and is likely to remain so for the next 50 years or so...
  • halik, on 10/12/2007, -1/+2you've completly missed everything...
  • Modulo, on 10/12/2007, -0/+1additionally, for future reference, reverse engineering a gun does not give you a bullet-sucking-out dewounding machine, reverse engineering a car crusher does not give you a car uncrusher and reverse engineering a light bulb does not give you a dark bulb. If you think I sound condescending for saying this, understand that this is actually exactly how stupid you sound when you talk about reverse engineering this system.
  • Spetz, on 10/12/2007, -0/+1All you need to know is that quantum physics is weird and wacky. ;-) And yes, trying to measure the 'spin' does disrupt it.
  • Modulo, on 10/12/2007, -0/+1let me attempt to clear this up for some of the "this is going to get cracked by reverse engineering" people. this is a system for totally uncrackable encryption. Reverse engineering the system will not allow you to decrypt the encryption, it will give you your own system for totally uncrackable encryption.

    The odds of someone being able to crack this encryption would be the same as the odds of me placing an apple on a counter, and having someone figure out a way to grab the apple from the counter and eat it while still leaving that apple on the counter.

    Is.. any of this.. getting through to you?
  • wthnow, on 10/12/2007, -0/+1coming soon: unhackable video game consoles
  • Vlatro, on 10/12/2007, -1/+1I don't doubt this as being possible, but I wonder who would ever need it. Current encryption schemes are extremely good. Brute-force attacks are a complete waste of time. Interception is logistically difficult. The types of computers that would use this I would imagine have some security measures in place to prevent eavesdropping anyway. The weak link in modern encryption is not the technology, it's the user. Personally I'd be more impressed with better interfaces and automation for the more advanced consumer encryption programs. Eliminate (or at very least reduce) the opportunity for user error, laziness, and general ignorance in encryption, and we'll reach a much higher level of security. I don't foresee foreign governments, terrorists or hackers posing much of a threat to the organizations who would use this technology. I can easily imagine some half-wit secretary at the pentagon e-mailing confidential data unencrypted to the bosses laptop at home. I can see corporate/industrial secrets being leaked because some jackass has schematics and/or legal documents stored on a computer that's loaded with crap-ware, keylogers, sniffers, Trojans, and icons on the desktop offering "100+ free music downloads that guaranteed to add 3 inches to his penis and make him rich in his spare time".

    The Heisenberg Uncertainty Principal
    Works theoretically but exact interpretation is widely disputed and often debated.

    People Do Stupid Things Theory
    The one and Only universal constant.
  • Chovee, on 10/12/2007, -0/+0 Ive dont that a hundred times today. I dont know jack about quantum theory, encription, or both of them together.
    Watching myself get dugg down is an easier method of higher education than learing Im dumb in school.
  • vuke69, on 10/12/2007, -1/+1Any encryption scheme worth using is still secure if reverse engineered, or just plain an open standard to begin with.

    You can understand AES for example, inside and out. And still be no closer to being able to break any given message, or all messages in aggregate.
  • chicken101, on 10/12/2007, -7/+7The Heisenberg Uncertainty principle is certainly my favorite principle.
  • Chovee, on 10/12/2007, -0/+0Good Point. I dont think this is even close to practical. And yes, I understand that the whole purpose is to monitor a man in the middle atack. I just ment to say that as far as encrytion goes. We dont need to worry about anything new.
  • evgen, on 10/12/2007, -0/+0While a quantum computer is expected to be able to crack most pubic-key methods in polynomial time using Shorr's algorithm (taking out systems based on the difficulty of factoring composites of large primes of of finding discrete logs), it is not going to take out symmetric encryption schemes unless an unlikely relationship between NP and BQP complexity is found to exist. I believe that it turns the effective key size n (e.g. 128 for standard AES) into sqrt(n) so much larger symmetric keys will be needed, but the basic algorithms will still be strong.
  • mailbox125, on 10/12/2007, -1/+1am i missing something here? this does not seem like a anything new, quantum encryption was first proposed in the 1970s and one time pads have been around even longer.
  • Chovee, on 10/12/2007, -0/+0Oh! I still think Mathamatic ignorance will outsmart the NSA.
  • tybris, on 10/12/2007, -1/+1"Scientists have created an unbreakable cypher through the use of quantum physics."
    NO. One-time pad is the unbreakable cypher (and very old), quantum encryption allows you distribute it 'safely'.
  • Chovee, on 10/12/2007, -1/+1How does "reverse engineering or whatever want to call it" break encryption?
  • Chovee, on 10/12/2007, -0/+0Thats argueable. There will always be astronmically high Prime Factorization. Imagine Trying to find the Factors of two primes a googolplex long. It would still be tough to get.
  • OsiVert, on 10/12/2007, -0/+0I think it's more on the lines that you know if someone tried to intercept your message since the pad has changed. But what I don't get is that this has to be used with light, correct? What happens when the signal over the internet gets transferred from fiber to say, DSL. Doesn't the light inherently get changed when it is passed through fiber stations anyways?
  • Fetching more discussions...
    Show 51 - 68 of 68 discussions

  • Add a Comment

    Login or register to comment!
    Or, sign-in super fast with your Facebook account ...

© Digg Inc. 2009 — Content created and posted by Digg users is dedicated to the public domain | Terms of Use Updated | Privacy Policy
DIGG, DIGG IT, DUGG, DIGG THIS, Digg graphics, logos, designs, page headers, button icons, scripts, and other service names are the trademarks of Digg Inc.