Re: Quantum Cryptography can not work

Joseph Ashwood wrote:
You cannot use an ordinary router for this link because there is no physical link in an ordinary router, you MUST have a physical link, routing this link is just like routing in the old telephone system, it would have to be significantly faster at switching though to be of any usefulness.

Okay, I didn't understand what you were talking about before, but I do now. When I said link I was talking about a coax cable (or whatever) physically running between two machines.

Ben Rudiak-Gould wrote:
That's equally true for a classical router, if you use optical delay lines.

Which is exactly what Quantum Cryptography requires, an optical delay.

No, it doesn't. (So neener.)

Classical only requires reproducing an approximation of the signal, quantum requires the exact signal, it cannot be duplicated.

But it can be transferred to a different system; there's nothing in quantum mechanics that forbids a transition like

| photon X > | electron up > ==> | photon X > | electron up >
| photon Y > | electron up > ==> | photon X > | electron dn >

or its inverse. By unitarity this implies

( alpha | photon X > + beta | photon Y > ) | electron up >
==> | photon X > ( alpha | electron up > + beta | electron dn > )

After this transition the photon's state is no longer correlated with the qubit state, so you can safely discard it into the environment. When it comes time to send the qubit onwards, you create a new X-polarized photon and apply the inverse transformation, then discard the electron. I don't know anything about the feasibility of engineering a transition like this, but I assure you that it works in theory.

-- Ben
.