May 3rd, 2005 04:17 PM
It is not a matter of technology it would be a matter of changing our understaning of the fundamental laws of nature and physics. You cannot view the state of a photon without changing the state of that photon. It can't be done. If you want we can get into the mathmatics behind it, but I barely understand a lot of it, so I don't think I can do it justice and it requires a pretty good knowledge of higher level calculus. Suffice it to say that it is not a matter of finding some neat new technology, it would require a monumental change in our understanding of the universe similiar to what Einstein and Newton did. Is it possible that our understanding could change this much? Yes, but I doubt it will happen. We actually understand the behavior of light, specifically photons, pretty well. The changes in our understanding of physics and the laws of nature will be in the weak force, the strong force, the eletromagnetic and gravitional forces and how those forces interact. That is where most research into quantum mechanics and string theory is taking us today.
So with current technology it is unhackable...Im sure wireless internet was thought impossible at one time also. My point was, I dont like people who say things like this are impossible, when in reality it will just take some time and money.
Like I said, it could happen, but most of the leading physicists in the world don't know how it would be possible. I only say it is possible because most of the physics behind this type of encryption is just theory with no practical way of confirming a lot of this. Our understanding and the mathmatics behind quantum physics is way ahead of what we can practical test. So what we know could definitely change.
Although, this claim that it can't be broken isn't marketting hype dreamed up by a ad guy that has no understanding of the technology. So it is a lot different than the CEO of Oracle saying his product is unbreakable.
Yes they could. But in quantum encryption if it is detected that someone else is listening in on the conversation the one time pad is recalculated and the data is resent. This will continue to happen until the transfer of data takes place with no one else listening into the conversation. If you can only detect one photon of the transfer before the keys changes you have no possibility of ever determining the entire key, or even what the encrypted data is.
One time pads can be broken, IF you want to spend 10-12 years working on one
May 3rd, 2005 11:40 PM
The theory behind quantum encryption is that if you listen or observe the conversation(intercepting the key), you change it, and the key is disregarded as a new one is generated.
This means that MITM attack would be pointless, but what about keys stored on the machine, or weak key generating algorithms ? Those could probably be hacked.
Trying to see in the box would be pointless, also. It is only 1 photon. Sent in 1 direction. It is not a blinking light bulb for everyone to see.
I think when they wrote that article, they meant 'hacked' as in intercepted. I don't think they meant totally proven. Just my $0.02
/edit ahh, mohaughn beat me to the 'post button' /edit
Also, ya gotta take in quantum computing. One time pads could be broken a lot quicker on a quantum computer than todays computers...
The fool doth think he is wise, but the wiseman knows himself to be a fool - Good Ole Bill Shakespeare
May 4th, 2005 01:54 AM
When you intercept a photon, and it changes (or you send your own and stop the original one), how do you know what the value was in the first place? Especially if there is only 1 and there is nothing to compare it against?
That is the problem that is being worked out via some sort of QKE protocol. Basically the system, to be 100% secure, must have NO redundancy. The problem is that running it like this makes it impossible to detect interceptions/modifications. From what I understand there is a tiny bit of redundancy in the protocol and a special way of discarding an agreed upon pair of bits. But as far as what I've read so far on a few protocols, none of this takes place over the quantum channel (yet).
Quantum Key Exchange is still vulnerable to Man-In-The-Middle Attacks. It's just monitarily expensive and computationally expensive to pull it off without being caught. Especially if you want to provide your own keys in the exchange and don't have the private key pairs of the victims.
I'd recommend learning about QKE and what issues are being raised/solved through something other than articles (because they're grossly over-simplified and take the "unhackable" approach to it without really telling you anything worthwhile), but finding the proposed protocols to solve problems in the system is difficult to come by. (I haven't managed to find the site I thought I bookmarked that looked to have some useful information...)
May 4th, 2005 07:03 AM
Details about the working of that device are not known to me, so
I won't comment on that. However, I agree with the fact, that there
are methods, which can be (mathematically) proven to be "unbreakable".
But one should not forget that a specific technology consists of several
parts, starting with hardware, software, ..., and the human element in the
end. Therefore, there might be flaws, which can be exploited, but still
the method itself could be "unbreakable".
For example, but in a different context than the thread started, but it
has been mentioned ...
the one-time pad
The one-time pad, properly used, is an "unbreakable" encryption scheme,
if properly created and used. If people create one-time pads with true
randomness (or "good enough" randomness (cosmic rays)), if people use
a one-time pad only once, and if the one-time pad itself is not visible to
an interceptor, then the one-time pad is "unbreakable".
A few people here suggest that one can break one-time pad encrypted
messages? I am wondering how. Would the message you extract the message
the sender has written? Could one of you elaborate? For example
starting with the definition of the one-time pad, such that we compare
apples with apples.
If the only tool you have is a hammer, you tend to see every problem as a nail.
(Abraham Maslow, Psychologist, 1908-70)
May 5th, 2005 05:38 AM
Here is some information:
Simplified QKE (Apparently something from Wikipedia since many sites have it) - http://www.absoluteastronomy.com/enc...yptography.htm
"Quantum Key Distribution over 67 km with a plug & play system" - http://xxx.lanl.gov/abs/quant-ph/0203118
The BB84 Protocol - http://www.cki.au.dk/experiment/qryp.../bb84prot.html
It seems each time I read more the information just floods in and changes what I thought earlier... (Though I'm correct in that some important parts of QKE -- at least in the protocols commonly used such as BB84, etc -- still happen over regular networks)
There is a large combination of using regular networks in QKE along with a quantum network. Since only a key is being generated via the network, you can exchange keys across both parts of the network until Alice and Bob agree on something (anything).
What apparently happens is that a random stream of photons are sent between Alice and Bob (Key Sender -> Receiver), and when something is detected (or not -- this part is random) Bob tells Alice (Key Receiver -> Sender) "I got something" or "Nothing" over a public channel while recording what was received.
A lot of data is discarded anyways since the phase of the photon emitter/detector will not always be lined up since the phase is set randomly by Bob and Alice seperately. It does no good to break into the public channel and say "I got something" while pretending to be Bob because chances are you wouldn't be able to get the key yourself. In a one time pad, you either have the whole key, or you're lost.
If you hop in the middle (MITM) though, you need to break into the Quantum network & Public network and work with Alice until you can properly negotiate a key. If you can make sure Bob never realizes he isn't talking to Alice (or Alice doesn't know its to you), you've succeeded in working out a key (and hopefully you discarded the correct bits). Then you just need to get the encrypted copy and decrypt that with the key you worked out while Bob is wondering why he can't decrypt it. Pulling this off is very difficult, though someone might do it "in-house" as an exercise and get published as breaking the previously unbroken network and spawning a whole new round of possible v impossible.
In the end though, the information is secured by the One-Time-Pad. The attacker doesn't know which solution is the correct one, and nearly every solution is possible. As long as the attacker doesn't know the key 100% to a one-time-pad encrypted chunk of data, they can't really determine if their solution is correct or not, and the information is considered secure... QKE is really just determing a key to encrypt something with so that nobody else knows what it [the key] is. As long as a bad person doesn't get the whole key through it, it is secure. It doesn't look like BB84 leaks much useful information, though chances are someone will manage to pull off figuring out a QKE generated key in their own lab...