Quantum Entanglement and FTL Information transfer

[FreshBranes]

First post.

There's something I don't quite understand about quantum entanglement and I'm hoping someone here can put me right. When I first heard about the Aspect experiment, I thought that it had to open the way forward for faster-than-light information transfer as follows:

Bob is on Earth and Alice is on Saturn. On Earth, Bob sets up his equipment to entangle photons, of which one half get sent to Alice. The other half stay on Earth but are effectively trapped by a cunning set of mirrors.

A few hours later, Bob checks his watch to ensure that the photons are about to arrive at Saturn. Bob then writes a message in binary to Alice. He dooes this by either measuring the x-axis spin of the first photon, or not observing it at all. He then repeats with the next photon etc.

On Saturn, Alice has a equipment setup for measuring the y-axis spin of the photons she's received. Now, either she can make a measurement or she can't - in the exact same sequence that Bob has been observing / not observing on Earth. So effectively she's receiving Bob's binary communication. And rather than travel at light speed, hasn't the binary code traveled instantaneously??

Can anyone shed some light on this?

[Don't get me wrong - I don't think I've discovered FTL information transfer or anything, I just want to correct my deficiency in understanding quantum entanglement and relatvistic effects?]

 

[selfAdjoint]

From your description, it didn't look like Bob got his particles with any particular orientation, so we have to assume the pairs were uniformly distributed across all angles. So Bob measures the spin of on particle in each pair around the x-axis, and I presume gets the standard three streams out, corresponding to +1, 0, and -1. Bob modulates the beam by turning the meansurement on and off. So a pulse of uniform distributions denotes bit 0 and a pulse of +1,0,-1 on the x-axis represents a bit 1.

Now Alice looks at her particles. They are also modulated, but in her case the different beams are all mixed together. How does she demodulate them? You say she measures the y-component. But a normal experiment shows that doing that on particles that have previously experienced an x-measurement should randomize the spins back to uniform distribution.

The key is that you get only one shot at the entangled state. Anything you do to configure it collapses the wavefunction and destroys the entanglement.

 

[slyboy]

In the experiment you describe, Bob encodes his information by either performing a measurement or not. Then you assume that Alice knows to either perform the mesurement or not in the same sequence that Bob did. Hence she already knew his information before the experiment started and can gain no information from it.

A slightly more sophisticated version of your experiment is as follows. Suppose Bob wishes to transmit a 0 to Alice. To do this he measures the x component of spin on his half of the EPR pairs until he gets spin up. If he wanted to transmit a 1 then he does the same thing except this time he waits for spin down.

No information is transmitted to Alice in this experiment, since any measurements she does on her EPR pairs have completely random outcomes and she does not know which one of the EPR pairs was used to encode the bit. However, if Bob tells Alice which EPR pair was used to encode the bit then she can find out what Bob's bit was by measuring the x-component of spin on her half of that pair. Notice that this transmission from Bob requires a signal, which cannot travel faster than the speed of light.

This is completely useless for faster-than-light communication, but notice that, if Alice and Bob really did share EPR pairs, no-one else would be able to determine Bob's bit. Thus, we have a kind of secure cryptography. The subtlety is that Alice and Bob must be certain that they really did share EPR pairs for this to work, i.e. that no-one tampered with the EPR particle whilst it was on its way to Alice. This is subtle because there is no measurement on a single pair that Alice or Bob could make to confirm this. However, they can get around this by selecting a random sample of their pairs and doing a Bell-type experiment on them. If they violate the Bell inequalities by a large enough amount, then one can show that this sort of cryptography really is secure.