Posted on 04/19/2002 6:55:52 PM PDT by Straight Vermonter
When I first heard about the encryption technique developed by Dr. Richard Hughes, it sounded like science fiction. After he explained it to me in detail, it still sounded like science fiction.
Imagine, if you will, a means of delivering encryption keys that is so secure that it's impossible to break because doing so would violate the laws of physics. In other words, the delivery method is so secure, it's protected by the very fabric of the universe.
IF THAT DOESN'T get your attention, think about this: What Dr. Hughes is working with is a way to encode information on individual photons. He then sends these encoded photons to a receiver that can measure their characteristics and determine from those characteristics the data that the encoding represents. That's right: He's imprinting information on individual subatomic particles.
What makes this so secure is that the information can be encoded in several ways. If someone were to intercept these photons, only one of the possible encoding methods could be seen. If that method didn't contain the needed information, the eavesdropper couldn't then look at the information that was encoded differently.
So why couldn't a hacker occasionally determine the encoding method by trial and error? To do so would be pointless, because hackers could never be sure they were seeing the real data, and it would violate the Heisenberg Uncertainty Principle of quantum mechanics if the hacker tried to look at both types of polarization on the same particle. Add to these complications that the mere act of observing the photons in their path changes them, and that the change is immediately detectable by the recipient.
HERE'S HOW all of this works. Dr. Hughes, who works in the physics division of the Los Alamos National Laboratory, creates photons using a very attenuated laser. He's able to choose the polarization of each photon so that, for example, a vertical polarization represents a 0 and horizontal polarization a 1. Or, you can use opposite diagonal polarizations.
The order of the polarization is varied randomly, but the sender has already sent the sequence to the receiver. That way, the receiver doesn't have to look at a bunch of other photons in search of the information; it just looks for the information coming from the transmitter.
The beauty of the method is that even if some hackers knew the order as well, they couldn't gather the information without being detected. Why? Again, Heisenberg: You cannot observe a subatomic particle without changing it. When particles that have been observed arrive at the receiver, their bit error rate is very high, alerting the receiver that the data stream is being observed. As a result, the encoding sequence can be changed immediately, regaining the security of the transmission.
NICE THEORY, RIGHT? But using polarized photons for encryption is more than just a theory. In fact, it's already working in at least two test installations: one is run by BBN Networks and Harvard University, the other by the Army and Navy Research Laboratories near Washington, D.C. Both installations transmit the photons over optical fiber. But transmitting photons over optical fiber is limited to about 70 km, because using repeaters to strengthen the signal would introduce anomalies similar to those of an eavesdropper. Dr. Hughes thinks the delivery method needs to go beyond that.
"A much more compelling application is by transmitting through the atmosphere," Dr. Hughes explains, adding that he's been able to do single-photon communications through the air in daylight with an acceptable bit error rate. No need to lay all that fiber-optic cable anymore. "The trick," he says, "is to find the single photon against the background." It sounds like a feat requiring equipment only NASA could afford--in fact, transmitting encryption keys to satellites is being tested--but Dr. Hughes says he's mastered "free-space quantum cryptography" using commercial off-the-shelf components.
Even better, you can buy this now. A Swiss company named id Quantique is already selling a device that performs over fiber networks what it calls quantum key distribution. While the wireless optical version Dr. Hughes is testing isn't available now, he says that at least two companies are working on commercial versions.
Once these products are available, you can be confident that your communications are secure. Unless someone starts messing around with the universe, anyway.
Will the laws of quantum physics end the theft of transmitted data once and for all? Why or why not? TalkBack to me!
Read more by Wayne Rash at ZDNet's Tech Update.
As a former USAF crypto guy (green door, lots of locked fire safes, and all that), I can say with some degree of verity that the military's encryption techniques in use in the 60's and 70's were impossible to break. Nearly. Almost. Probably. We hoped so, anyway.
The trick was to change the key often, so that the old key was rendered useless before it could be broken and any useful info extracted. No doubt, state of the art has made giant leaps forward since then, but the principle is the same - just a different way of changing the keys.
Let's suppose a hacker does try. If he tries the wrong method, he gets junk. If he tries the right method, he gets clear text -- "Now is the time for all good men..." In other words, he'll know when he gets the right data because he can read it.
There is only one truly unbreakable encryption method. The one time pad. And that's only unbreakable if you only use it once. And if the pad was truly random. Use the same pad twice, and people have a chance to break it.
With a one time pad, if you get clear text, you don't know if it's the right clear text. You can never know unless you have the one time pad to decipher the message.
Plus, this method isn't usable in the real world. As pointed out in the passage quoted above, attempts to read the message changes it. In other words, I don't have to read your message. All I have to do is intercept it, do whatever to the particles, and you won't be able to read it either. If the message is important enough, you'll have to resort to another method of encryption. That, unless you use one time pads, can be broken.
Shhhh! Traffic analysis is classified.
As another person who lived behind the green door, I'd say that you hit the nail right on the head. The only way to make the thing unbreakable is to change the key with every message.
In the US, privacy is a crime. Funny, the "right to privacy" allows women to have an abortion -- an essentially public act since it involves doctors, nurses, clinics, insurance companies, etc. But real privacy -- in the sense of keeping government from prying into our lives is non-existent and even illegal.
Quantum key crypto is just a way of sending keys to two locations that guarantee that no-one can intercept it without violating Heisenburg. Think of it as one time pad distribution that cannot be intercepted.
The message is still as vulnerable as the algorythem/key length combination that are used with the key. One time pads traditionally used one key to encode one character and such were not vulnerable to computational attacks.
Well designed encryption used correctly is already unbreakable. Also you can generate quite good pseudorandom pads which combined with a good encryption NEVER will be broken by honest methods. But the old rubber hose method is still there.
Disclaimer: Opinions posted on Free Republic are those of the individual posters and do not necessarily represent the opinion of Free Republic or its management. All materials posted herein are protected by copyright law and the exemption for fair use of copyrighted works.