A New Unbreakable Code System

[Mr. Anderson]

http://www.cnn.com/2005/TECH/science/05/03/australia.cryptology.reut/index.html

Very cool stuff and the fact that they can get a single photon beam of light out of it is amazing. But they don't mention more about how you'd implement the system or what it takes to *read* a single photon.

D

[robbieduncan]

From what I understood this is not a code system, more a transmission method. It does beg the question what happens if the read end misses a photon? I assume the transmission protocol must have quite a lot of error checking built in...

[Kyle?]

This system would seem to be unstable over relatively long distances, since some slight leakage is expected when transmitting over fiber. This would likely require the signal to be interrupted more frequently to check for errors and to realign the signal.

[jsw]

Cool. According to this article (http://www.commsdesign.com/showArticle.jhtml?articleID=29106041), this new single-photon method prevents the man-in-the-middle attack developed for prior quantum-entanglement-based schemes.

Of course, quantum entanglement (in its original, multi-photon form) was at first considered unhackable, so one must wonder how long this new method will be seen as such.

[Mr. Anderson]

Of course, quantum entanglement (in its original, multi-photon form) was at first considered unhackable, so one must wonder how long this new method will be seen as such.

Until someone learns how to just observe the photon without removing it from the stream :D

D

[Chappers]

From what I understood this is not a code system, more a transmission method. It does beg the question what happens if the read end misses a photon? I assume the transmission protocol must have quite a lot of error checking built in...

Don't worry - Microsoft are working on it.

[Wes]

Until someone learns how to just observe the photon without removing it from the stream :D

D

Hmm isn't that very hard to do as a result of Heisenberg's uncertainty principle? The process of detecting such a tiny mass (http://math.ucr.edu/home/baez/physics/ParticleAndNuclear/photon_mass.html) is going to deflect it to a huge degree, hence alerting to a hacker intercepting the signal.

[fedora]

Don't worry - Microsoft are working on it.

No apple are - OS X Schrödinger's cat, for release on the first desktop quantum computer. :D

[clayj]

Hmm isn't that very hard to do as a result of Heisenberg's uncertainty principle? The process of detecting such a tiny mass (http://math.ucr.edu/home/baez/physics/ParticleAndNuclear/photon_mass.html) is going to deflect it to a huge degree, hence alerting to a hacker intercepting the signal.It's IMPOSSIBLE to do, unless Heisenberg was wrong.

[fedora]

If anyone want to learn more about quantum computing/physics come and do a search on this forum I post on: www.physicsforums.com this topic and it's link too LQG have been discuss in depth.

[Wes]

It's IMPOSSIBLE to do, unless Heisenberg was wrong.

Or unless we find smaller particles to detect photons with...

[fedora]

Or unless we find smaller particles to detect photons with...


What do you mean "smaller" photons have no dimensions? only mass as a product of velocity.

[clayj]

What do you mean "smaller" photons have no dimensions? only mass as a product of velocity.They still have size, though, right? A point particle (dimensionless) with mass would have infinite density (weight per unit volume), and that doesn't make sense.

I think he meant trying to use quarks or even strings to detect and "read" a photon's state without altering it. Of course, if you could do this, the resulting apparatus would be so large and require so much energy to operate that everyone would know what you were up to.

[admanimal]

No apple are - OS X Schrödinger's cat, for release on the first desktop quantum computer. :D

The retail box contains a single dvd that is both blank and contains the final version...you don't know which you get until you open the box.

[cr2sh]

God I really hate threads like this. Everyone has to whip out there college physics notes and try to pass judgement.

Aren't there insanely cheaper and simpler "unbreakable" codes?

[ShaunPriest]

Anyone in here read Dan Brown's 'Digital Fortress'. It's about a code that cannot be broken - I'm in the middle of reading it!

[clayj]

God I really hate threads like this. Everyone has to whip out there college physics notes and try to pass judgement.

Aren't there insanely cheaper and simpler "unbreakable" codes?Yup. The good ol' one-time pad. Use the cypher once, then throw it away. The people at both ends of the communication have the same pad. Since the cypher is NEVER reused, a pattern can't be established by anyone who's monitoring the communication, and therefore they can never decypher it.

[admanimal]

God I really hate threads like this. Everyone has to whip out there college physics notes and try to pass judgement.

Aren't there insanely cheaper and simpler "unbreakable" codes?

Well since we're getting all theoretical here, it's worth pointing out that the varying levels of "unbreakableness" achieved by modern (non-quantum) cryptography are all based on the assumption that P does not equal NP, i.e. that there are certain reasonable things that we would like to compute efficiently, but can't seem to (the prime factorization of an integer being the most relevant). The disturbing thing is that it is unknown whether it is truly impossible to compute these things efficiently or whether we just haven't figured out the trick to it yet. If someone does find the trick, then most modern encryption techniques would fall apart.

So I don't get attacked by any theoreticians hanging out here, I should point out that finding a polynomial algorithm to solve the factoring problem (the NP problem most relevant to encrpytion) does not prove that P=NP, since it is not NP-complete.

[MongoTheGeek]

Until someone learns how to just observe the photon without removing it from the stream :D

D

Or goes through the trouble of replacing it...

[fedora]

They still have size, though, right? A point particle (dimensionless) with mass would have infinite density (weight per unit volume), and that doesn't make sense.

I think he meant trying to use quarks or even strings to detect and "read" a photon's state without altering it. Of course, if you could do this, the resulting apparatus would be so large and require so much energy to operate that everyone would know what you were up to.

I know this is a bit off topic,

You are correct, it's correct to think that photons do have "size". I said they did not because the physics gets difficult. A photon has a perfectly well-defined wavelength only when it's in a momentum eigenstate, i.e. when it has a perfectly well-defined momentum (and energy). This never happens. A photon is always in a superposition of momentum eigenstates:

http://img188.echo.cx/img188/9291/picture17xs.png

The only quantity that we might want to call the "size" of the photon is the width of the Fourier transform of the momentum-space wave function, f, i.e. the uncertainty in the photon's position. This uncertainty could be anything between zero and infinity. Since it can be arbitrarily close to zero, it makes sense to call the photon a "point particle".

However, if we assume that the uncertainty in momentum is proportional to the magnitude of the momentum (which is the only thing we can assume if we know nothing about the state), the uncertainty in position is proportional to Planck's constant divided by p (the magnitude of the momentum). Since p is inversely proportional to the wavelength, the uncertainty in position is proportional to the wavelength.

So it makes sense to think of the wavelength as the "size" of the photon (or at least as something proportional to it). This may seem strange, but it is at least consistent with e.g. the fact that microwaves (with wavelengths of order 1 cm) won't go through a metal net with millimeter-sized holes (like the net that covers the window of your microwave oven), but they will go through a net with much larger holes.

So photons do have "size" but it is not size as in a lump.

[yellow]

Australian scientists believe they have developed an unbreakable information code to stop hackers, using a diamond, a kitchen microwave oven and an optical fiber.

I wonder what they would have found if they'd added a bag of microwave popcorn to the mix.

[robbieduncan]

Anyone in here read Dan Brown's 'Digital Fortress'. It's about a code that cannot be broken - I'm in the middle of reading it!

Yes I've read and. And I don't think I'll ever forgive Dan Brown for writing such a terrible book. Probably the worst thing I've read in the last 2 years :mad:

[Nickygoat]

Yup. The good ol' one-time pad. Use the cypher once, then throw it away. The people at both ends of the communication have the same pad. Since the cypher is NEVER reused, a pattern can't be established by anyone who's monitoring the communication, and therefore they can never decypher it.
They are unbreakable on a depth of one (used only once) but are unsuitable for mass market systems. They are also vulnerable to social engineering ie. capturing one of the one time pads.

[cr2sh]

Yup. The good ol' one-time pad. Use the cypher once, then throw it away. The people at both ends of the communication have the same pad. Since the cypher is NEVER reused, a pattern can't be established by anyone who's monitoring the communication, and therefore they can never decypher it.

Exactly the system I was thinking of... the system that killed the Rosenbergs.

http://www.math.ucsd.edu/~crypto/Projects/KyleCai/spiesandcrypto_files/one%2520time%2520pad.jpg

[clayj]

They are unbreakable on a depth of one (used only once) but are unsuitable for mass market systems. They are also vulnerable to social engineering ie. capturing one of the one time pads.Well, that's true for ANY system. You have to keep both ends of the communication secure.

Tom Clancy described a (fictitious) one-time pad system called Mercury in one of his novels... rather than a physical pad of paper, the "pad" existed on a CD-ROM and was generated using atmospheric radio noise, so that it was truly random. (If you let a person come up with the cypher, they'll unknowingly apply a pattern to it given enough time.) Two copies of the CD were made... one for the base, and one for the remote location. All transmissions were encrypted/decrypted using the contents of the CD, which was then destroyed once all of the bytes on it were used.

Using a dual-layer DVD (8.5 GB of random bytes) at both ends, you could encrypt/decrypt a HELL of a lot of e-mail...

[Mr. Anderson]

I just mentioned the reading the photon without effecting it more as something that could happen if someone figures out how, not based on current technology or current theories.

The one thing that seems to happen to stories like these is that someone comes along eventually that *solves* the problem. It will actually be more of an issue if we see this 1 photon beam actually being used outside of a lab.

D

[jsw]

Of course, no matter what the encryption, paying someone a year's salary will likely get you the info.

[savar]

I wonder what they would have found if they'd added a bag of microwave popcorn to the mix.

Not many people make me laugh on MR, but you just cracked me up. I'm glad you like VB, too. One of my friends interned at Adult Swim for a few summers, so I got to watch the VB pilots before they even signed them.

[savar]

It's IMPOSSIBLE to do, unless Heisenberg was wrong.

Can somebody explain to me why you still couldn't do "man in the middle"? Just with the same receiving and transmitting equipment that they are using, read the signal in and save it for your own purposes, then transmit the same signal out on the other end.

[clayj]

Can somebody explain to me why you still couldn't do "man in the middle"? Just with the same receiving and transmitting equipment that they are using, read the signal in and save it for your own purposes, then transmit the same signal out on the other end.EDIT: OK, better explanation... say I have a system like this set up, and I am sending you a message. The system is designed to send one photon every X seconds (where X is some small extremely small time interval, like a billionth of a second). So here I am, sending photons... ping, ping, ping, ping... and you're receiving them.

If someone intercepts my photons, the timing will be off because the speed of light is constant and because of the time necessary to intercept the signal and resend a new photon in your direction... so instead of ping, ping, ping, ping, you'd get ping, ping... pause... ping, ping. The time gap would be the dead giveaway that the signal had been intercepted.

[jared_kipe]

EDIT: OK, better explanation... say I have a system like this set up, and I am sending you a message. The system is designed to send one photon every X seconds (where X is some small extremely small time interval, like a billionth of a second). So here I am, sending photons... ping, ping, ping, ping... and you're receiving them.

If someone intercepts my photons, the timing will be off because the speed of light is constant and because of the time necessary to intercept the signal and resend a new photon in your direction... so instead of ping, ping, ping, ping, you'd get ping, ping... pause... ping, ping. The time gap would be the dead giveaway that the signal had been intercepted.

Even easier, when you go to read a photon you have to guess as to what polarization axis the transmitted signals are in. If you guess wrong, then you don't get the right message. The people on the end us SOME of the data to publicly confer that they got the right axis guess. A man in the middle would fudge this up. MiM would have to make a guess, but then all he could do is send what HE got back out, it would be very easy to get, not only the time delay, but when the actual recievers decied to "compare notes" they would have to find differences and throw out it all away. Remember that both the message, and the decryption code are never sent through the same means. Whatever the MiM got, he can't use anymore.