Follow Slashdot blog updates by subscribing to our blog RSS feed

 



Forgot your password?
typodupeerror
Check out the new SourceForge HTML5 internet speed test! No Flash necessary and runs on all devices. ×

Growing Diamonds for Better Information Security 113

hip2b2 writes "NetworkWorld is running an article that describes how a University of Melbourne research group is developing technology to make fiber optics communications more secure. The technology is based on Quantum Cryptography principles and requires than absolutely only one photon gets sent at any given time. Today, fiber optic systems do not send one photon at a time. They only approximate it. This makes current systems unsuitable for their secure communications technology. Therefore, the group uses artificially grown diamonds to achieve this."
This discussion has been archived. No new comments can be posted.

Growing Diamonds for Better Information Security

Comments Filter:
  • ...But is it SO hard to proof read this stuff.IANAGN (i am not a grammar nazi) but I got up 5 mins. ago and saw that from across the room :)
    • ScuttleMonkey is running an article that describes how hip2b2 is automatically generating articles to get better slashdot coverage. The technology is based on bayesian generation and requires that articles are not read in depth, but merely scanned. Today, articles are not read. Slashdot readers only skim them. This makes random generation suitable for getting articles posted. Now I will mention diamonds.

      Congratulations, you have very nearly passed the Turing Test.

  • by layer3switch ( 783864 ) on Sunday May 07, 2006 @06:40AM (#15280722)
    Quantum Cryptography Field will be soon swarmed with females. INGENIUS! University of Melbourne research group just came up with an answer for the problem on this total sausage party we have going on with CS department.
  • by dbIII ( 701233 ) on Sunday May 07, 2006 @06:41AM (#15280723)
    Grown diamonds are literally vaporware - but chemical vapor deposition is the interesting and relatively cheap way to do it. The old cheap way to make artificial diamonds was to blow things up (DuPont method), but the optical properties were no good.
    • Not all grown diamonds are vaporware in the literal sense. - Don't remember which university it was but they actully turned peanut butter into a diamond using a relatively easy process. Charbroil the hell out peanut butter until it's been reduced to carbon ash, grind the carbon ash up to fine dust, apply immense amounts of heat and high-level atmospheric pressures to simulate the pressure of large amounts of land - wait a while. You now have a diamond - it's optical quality was HORRIBLE (horrible is actuall
      • http://www.wired.com/wired/archive/11.09/diamond_ p r.html [wired.com]

        That article talks about both the vapor process & some old Russian technology that squished the hell out of stuff to make diamonds.

        The part of TFA that interests me most is this

        This is achieved by "growing" diamonds, which are "usually cleaner" than the mined gems, in QCV's lab. The synthetic diamonds have a defect which is the source of the single photon.

        Defect?

        • Defect?
          http://en.wikipedia.org/wiki/Dislocation [wikipedia.org] If I remember correctly
        • That article talks about both the vapor process & some old Russian technology that squished the hell out of stuff to make diamonds.

          I've got a book called The Rise and Fall of Diamonds (IIRC) that talks about earyl attempts at artificial diamonds. The russians had an unprofitable device, but GE made the first profitable artificial diamond maker. It was a huge set of hydraulic rams that came to points; where the points came together, you made a diamond. IIRC they sold the technology to DeBeers.

        • You've got a good article, and so do the posters below me (maybe above me due to karma level?) but most of those methods have been made inefficient with the advent of thermal induction thru radiation (similar to the glass-top stoves we have today, I believe?) We just vaporize it, then subject the resulting carbon ash + impurities (I'd like to see how they filter out nothing but carbon from that,) to intense levels of atmospheric pressure (thanks to high-tensile strength alloys and to how thick they were bui
  • No popups (Score:3, Informative)

    by Kangburra ( 911213 ) on Sunday May 07, 2006 @06:42AM (#15280727)
    Here's the actual University of Melbourne [unimelb.edu.au] article from four days ago.
  • Diamond-based devices could be helping IT managers detect network snooping and prevent information theft The technology, based on quantum cryptography, uses a diamond to produce a single photon of light to stop information being intercepted

    Will it not increase DOS attacks, if the attacker's aim is not the information theft?

    • Re:Wait (Score:3, Insightful)

      by fabs64 ( 657132 )
      Presumably if they've managed to get access to the optical network and wanted to DoS it, they could just cut the fibre.
  • EXCELLENT! (Score:1, Funny)

    by Silmeria ( 972282 )
    Now I can buy some of these cables without my wife hounding me for justification.
  • Many teenagers working a crap supermarket job remember the number 1 rule: steal from the store, you're in trouble. Well with that in mind, it's safe to say the manager of this project would be packing a serious Uzi.
  • by Proudrooster ( 580120 ) on Sunday May 07, 2006 @07:38AM (#15280809) Homepage
    First-generation products will be for very secure transmission of secure datasets, like a bank's daily offsite backup, but could serve the commodity networking market in about 20 years, Huntington said. It's a low transfer rate but idea is not to send data [this way] but the encryption key so you don't need the same transfer rate. One of the consortium's goals is to enhance that as much as possible. If you can securely transfer the key you can transfer the rest of that data over a standard telco line, he said.

    So let me get this straight. The article implies: 1) I can build a secure fiber line between two points and to transfer a key, one photon at a time; and 2) once the key is transferred, I can then use standard telco lines. If I am going to the trouble to build a custom fiber optic network between two points that works with diamond lasers, why would I use telco lines? Conversely, if I don't build my own point to point fiber for key transmission then I run the risk of man-in-the-middle stealing my keys since the middle will have repeaters which can regenerate these 'secure photons'.
    I say to you, this makes no sense. Why not just put 52 keys on a thumb drive or CD (one for each week of the year) and send it via a secure courier and then use telco lines for transmission? This looks like yet another ruse to get research money under the guise of quantum cryptography.
    • Consider the world. Consider your picture of the world. If they are different, your picture is wrong.

      Money is flowing into quantum crypto because courier-based kex is insufficient. Also, QC is intrinsically point-to-point since there's no current way to reliably switch photons. This allows you to take two black boxes and connect them with a cable 20 miles long, and you're 100% guaranteed to be able to get information from A to B without anyone being able to find it out. Could be good for, say, teleconferenc
      • Consider the world. Consider your picture of the world. If they are different, your picture is wrong.

        Are you aiming for 'zen' or 'nonsensical' here? If I can "consider the world" in such a way that I am not really considering my own perception of it, don't I have access to the objective truth already? (Meaning that I would be basically omnipotent, and in no need of philosophical advice from random people on Slashdot.)
        • I was aiming to point out logical inconsistencies in a person's own subjective world view. How can you reconcile "my idea is great" with "they are not using my idea" other than to assume logical inconsistences (such as: a large rich company with excellent R&D didn't think of it) and/or unlikely assumptions (such as: I'm better than them, that's why they didn't hire me) ?
          • I was aiming to point out logical inconsistencies in a person's own subjective world view. How can you reconcile "my idea is great" with "they are not using my idea" other than to assume logical inconsistences (such as: a large rich company with excellent R&D didn't think of it) and/or unlikely assumptions (such as: I'm better than them, that's why they didn't hire me) ?

            I also have an issue with your black-and-white statements regarding a person's subjective world view.

            How does one reconcile "my id
      • I am more interested in "quantum entanglement" which Einstein referred to as "Spooky Action at a Distance." Assuming one could entangle particles, then manipulate and observe them, one could be theoretically build the ultimate secure communications device. Additionally, this device might possibly work at superluminal speeds. Imagine controlling a space probe on Mars without the 7 minute delay. Who knows what is possible, sub-space communication anyone? :)

        As for your "Consider the world" argument. I
        • It was more of an argument against people who are arrogant enough to have opinions like "my idea for a mars rover is better than nasa's, let's just stick xeons in it" or something. People who don't even consider the possibility that the entity they are criticising has already thought of whatever they propose and dismissed it.
    • The point of building/using a quantum channel (the fibre line) is to solve the key distribution problem, it cannot be used to send data. Why? Firstly in the protocol used for checking for eavesdroppers you end up discarding around 3/4 of the photons sent, with no way of predicting which ones, and secondly you really need to be sending random data to make it completly secure. The result is both parties end up with a random key, and you know with absolute certainty that no-one else has it. Compare with your "

      • Once you have your key though, the can use the Vernam cipher (one time pad [wikipedia.org]) which is provably unbreakable, to send the actual data over a standard telco line, copletly securely.

        No, the key for a one time pad is just as long as the data itself. So if the quantum network has too little bandwidth to send the whole message, it also has too little bandwidth to send the key for a one time pad.

        I am curious why the quantum net itself is necessarily slow? It can't just be that 3/4 of the photos

        • Yes, you can only send as much data as you have key (well securely anyway), but thats not the reason for only using the quantum channel (fibre) for key distribution. The point is that you need to send a completly random string of data down your channel, and then completly at random discard about 3/4 of this. So your left with a shared random key, which you can use as a one time pad to send your actual message.

          The whole thing is slow (at the moment) as all the technology is very experimental and you need to

          • But the decision about what is and isn't discarded also has to be transmitted over your link, otherwise your one-time-pads won't match up, I would think?

            Besides, if you're going to transmit as much key as you have message, why use two different lines at all? Why not use some currently "secure" method over the inherently secure quantum line, and not have to send twice as much data?
        • You can build the "one time pad" over a few hours and then use it over a few seconds. Basicly, the key needs to be sent fast enough to meet the average demand but the link needs to handle the spikes. There are other constrants; such as if you send both links though the same line then you can suffer a man in the middle attack but if they are on seperate networks then it's much harder.
      • Uhm.. why couldn't one intercept that lone photon, send a copy to the attacker, then retransmit it at the other end ? How can the receiving end tell that its communication has been intercepted ? Data is data.. if I intercept your phone or data lines, but make sure the forwarded signal is identical to the original, you have no information upon which to detect my presence. It's a classic black box scenario.
      • by Anonymous Coward
        Quantum cryptography is invulnerable to observation, but it is still vulnerable to a man-in-the-middle attack in which the attacker pretends to be the destination. From wikipedia:

        Quantum cryptography is still vulnerable to a type of MITM where the interceptor (Eve) establishes herself as "Alice" to Bob, and as "Bob" to Alice. Then, Eve simply has to perform QC negotiations on both sides simultaneously, obtaining two different keys. Alice-side key is used to decrypt the incoming message, which is reencrypte
    • The actual encrypted message is sent using a one-time pad encryption scheme over standard telco lines. This is secure because a one-time pad can be totally unbreakable if you don't have the pad (the key). So this means that the real problem lies in key distribution. Using the method you suggest (courier), there's no way to detect a man-in-the-middle attack. However, due to the nature of quantum cryptography, it is invulnerable to a man-in-the-middle attack, because as soon as someone begins intercepting/det
  • As i understand it - the idea is to build a point to point connection between two boxes, and transfer the key really slowly, so that noone can eavesdrop ?
    If something can be read, and written - it can be copied. It might be harder using this technology, but as soon as it goes global - and the devices capable of generating a single photon impulse and reading an impulse like that are available (even if for a horrific price) the strategy goes to hell.

    If someone is capable of listening on a optic fiber in the p
    • As I understand it, but of course I am not a quantum crypto researcher, the idea is that this is secure because your premise of "If something can be read, and written - it can be copied." does not hold true.

      The idea is that with these quantum particles you are transmitting the data by means of the "spin" property of the particles, rather than simple on/off pulses. The key point is that by measuring the spin you affect it and change it completely, meaning that anyone at the other end will know, because a
    • Gee, too bad those silly boffins at Melb Uni don't have you there advising them.

      Public notice: If you see the word "Quantum" in a topic summary, and have no bloody clue, and are too lazy to research it; go right ahead and skip over it.

      http://en.wikipedia.org/wiki/Quantum_cryptography [wikipedia.org]

      In short, theoretically you cannot eavesdrop (successfully) on a link using quantum encryption. So that makes "practically" completely out of the question when someone has managed to get it down to single photon transmiss
    • by centie ( 911828 ) on Sunday May 07, 2006 @08:20AM (#15280889)
      As i understand it...

      No no and no. I'm not meaning to be harsh but everything you said is misunderstood. The point is not to achieve security by doing things really slowly, the point of quantum cryptograhpy (wikipedias quite good [wikipedia.org]) is that if anyone intercepts your photons/information, you know about it. So you can resend the information, using a differenet channel, whatever. It is very important in crypto to be able to guarantee that no-one else has your key.

      If something can be read, and written - it can be copied

      Entirely true in the classical, everyday world, and you'd think so on small scales (individual photons/atoms) too.. but actually wrong. Quantum states cannot be coppied (no cloning theorem [wikipedia.org]). This is where the security of quamtum cryptography lies. There's nothing to stop someone from eavesdropping on your fibre, but if they do intercept anything you know about it. The only way they can get information without you knowing is if you accidently send the information twice, ie two photons in a pulse instead of one. Thats where this research is useful, its anything but pointless.

      I don't see how transmitting single photons at a time as opposed to the millions used today would give a speed increase, the fastest quantum cryptography demonstrated so far achieved a rate of 500b/s, compared to 500Mb/s for normal fibre communication. It's only real purpose is cryptography.

      • >Quantum states cannot be coppied

        That's true and key to the whole idea of quantum key exchange. People rarely mention the implications.

        You can't conduct quantum key exchange through a repeater. The protocol is only useful for your friends and business partners who are within unboosted fiber range.

        Raising the question, why not put your key material on a 400G disk, put that in a tamper-evident container, and ship it via one of the armored car services that routinely handle shipments worth USD tens of milli
        • Raising the question, why not put your key material on a 400G disk, put that in a tamper-evident container, and ship it via one of the armored car services that routinely handle shipments worth USD tens of millions?

          Because even those get jacked from time to time. Yes, it's secure, but it's not totally secure. Remember, the point is to make sure that the key is never discovered. If you detect someone eavesdropping on your quantum key exchange, you scrap that exchange and repeat it until you get a clean, u
          • That's not why. There's a simple protocol for avoiding that problem. Send the key via armored vehicle, repeating as necessary, until it is successfully sent without interception. You can then send the encrypted message either via courier, or just over the open 'net, provided you are using sufficiently strong encryption. The reason they need to do this is that they need low-latency communications, for instance for wire transfers.
      • Perhaps I still don't get it, but didn't public key encryption solve this problem years ago?
        • In a way it has, temporarily. All public key encryption relies on the difficulty of factoring primes (or something very similar). It is not proveably secure (unlike quantum cryptography), and in fact no-one has even proved that factoring primes cannot be done efficiently classically, we just don't have a way yet. One of the main reasons for interest in quantum computing is its ability to factor primes (thus breaking public key encryption). So in a way quantum cryptography is a solution to an anticipated nex
    • If someone is capable of listening on a optic fiber in the present day - and im fairly certain there arent many people like that out there - whats to stop them from eavesdropping on a fiber such as this ?

      As another poster just said, it isn't about them not tapping the line, but rather that you instantly know if someone is listening in. Heck you could even automate it to shut off communication if someone taps the line.

      Of course we are talking about easy DoS attacks, but this application is for those who need
    • If something can be read, and written - it can be copied.

      Copying takes time. Any system like this would be carefully timed with an atomic clock on both sides so that a latency change would immediately be discovered. It has always been possible in a fiber system to detect an active man-in-the-middle by monitoring the latency.

      This prevents passive listening, where a portion of the beam is split off and monitored. If you're only sending a single photon, there isn't a portion to split.

  • Therefore, the group uses artificially grown diamonds to achieve this."

    Surprisingly, the new diamond cables are still cheaper than Monster Cable. :^)
  • I suppose you could detect it while it's happening and shut down.
  • I'll say that again and again, until people listen:

    Quantum criptography doesn't work!

    Well, it does work on the sense that if you have a secure channel, you can use it to validate another channel. It doesn't work on the "do something usefull" sense of the word.

    Now, growing diamonds are interesting, and may be usefull for lots of things. But not for quantum criptography, because QC is not usefull.

    End of Rant (EOR)

    • I'll say that again and again, until people listen:

      Quantum criptography doesn't work!

      It doesn't work on the "do something usefull" sense of the word.

      I guess people ask you again and again:

      What the hell are you talking about!

    • 1) Cryptography, and 2) Useful.
    • I have the... dubious... honour of knowing some of the researchers on this particular project. I don't claim to be a cryptographer (READ: Freak with less life than a normal computer tech ;)) but I had them take me through this topic anyway. Quantum Crypto does IMHO work. At least the theory of it does.

      It seems that a lot of people (in regards to the security of this) are missing the point. This particular project aims to secure the physical medium.

      Points of attack
      The exchange where the links are conne
      • Yes, it does what it is inteded to do. No, what it is intended to do is not useful. Let's go...

        If an attacker is able to intercept and change the messages of any end. Let's say that A wants to talk to B, and S is able to intercept and change the messages (man in the middle). Now, S can autenticate with A and B, and none of them will notice him.

        Mathematical cryptography avoids this problem by A knowing how to validade a message that only B can generate, and B knowing how to validate a message that only A c

        • "Garantee" is such a strong word that it doesn't even exist.

          Perhaps you mean guarantee?

          Sorry to be such a spelling Nazi, but since you were so assertive, I figured you might as well spell it right.
        • That means that to break a QC key exchange, S will need only to cut the fiber and link both ends to his computer.

          There are practical implications. Any kind of repeater I can think of, built using today's technology, like transistors and integrated circuits, will insert a very high delay.
          Suppose there is a 100Km quantum line between two points, and that the speed of light in the line is 200.000 Km/s, this will generate a very stable latency of 0.5 msec.
          So depending on the accuracy of measuring this la
  • It will mostly benefit the military. Everyone else will have to live with internet communications being monitored by a mix of FBI/NSA/Dept of Homeland Security.

    Nobody even cares to encrypt email... I believe the main obstacle to more secure communications is human, not technical.

  • I can't see how sending one photon at a time will make a system secure. Photons are not necessarily particles, they have wave properties too. So if particle domain analysis doesn't work, just use the wave domain. Have these guys never heard of the double slot experiment?
  • Dunno if the system will *work* or not, but I'm sure there'll be a lot of law-making bee_ess involved (terrorism / RIAA) that'll bring down the whole system.

    PS: Hey! I managed to get RIAA and terrorism together ;)
  • Check out the link from the Network World article to their

    Alpha Doggs Blog [networkworld.com] .

    This is described as "The future of networking as seen through the works of university and other labs"; it's the best name for a tech blog that I've seen in a while.

HELP!!!! I'm being held prisoner in /usr/games/lib!

Working...