Dismiss Notice
Join Physics Forums Today!
The friendliest, high quality science and math community on the planet! Everyone who loves science is here!

Faster than light information travel

  1. Nov 8, 2012 #1
    Faster than light information travel aparatus

    Ok, so I am not a physics pro, but I find it very interesting. And I just finished reading about Bell's theorem and I have a question.

    As I understand it, the laws are like this for entangled photons:
    1. If you measure the polarization of photon 1 at angle A, you know the polarization of photon 2 at angle A is the opposite.
    2. The polarization at angle A is reset at random if you measure any other angle, but otherwise stays the same

    So lets say that I build a device on earth that spits out entangled photon pairs rapidly and continuously. And lets say the device is configured such that we know the photons that the device emit from the left side will reach mars after 7 minutes. The photons that the device spits out from the right side on the other hand goes through a path of mirrors that takes exactly 6 minutes and 59 seconds to traverse.

    Now lets say that on mars we have another device that detects these photons and reads their A angle and interprets the result as bits like on a normal network.

    We start up the transmitter and after 6 minutes and 59 seconds we measure the A angle on earth. If we want the A angle to be up and it is already up, we simply keep the photon for 2 more seconds before letting it escape. If the angle is not what we want it to be, we measure angle B to reset and then measure angle A again and keep doing this until we have the preferred value for angle A. And lets say that our advanced machine can do this 1000 times in one second so that we have only a 1/2^1000 chance of not aligning our photon pair before the receiver reads it. If both mars and earth had a receiver and a transmitter, this noise and other noise could be mitigated with check summed packages such as it is done already on networks today.

    If such a device was possible, I am sure others would have thought about it before me, so where am I wrong?
    Last edited: Nov 9, 2012
  2. jcsd
  3. Nov 8, 2012 #2
    Re: Faster than light information travel aparatus

    The photons need to be compared......every time/one.

    The photon A (on earth) needs to be compared with photon B (on mars).
  4. Nov 9, 2012 #3
    Re: Faster than light information travel aparatus

    On Mars they would look at angle A and know that polarized means 1 and non polarized means zero. So why would they need to be compared?

    I appreciate your reply, but you didn't really explain anything.
  5. Nov 9, 2012 #4


    User Avatar
    Science Advisor
    Gold Member

    Re: Faster than light information travel aparatus

    Your rules are more or less correct. There is no question you can delay measurement of one photon of the pair until after the other. The issue here is that you must properly label the photons and consider what you are doing to each. So we have Alice on earth trying to send a message to Bob on Mars, and you want the message to be communicated in let's say 1 second (which would exceed the speed of light c). Alice measures her photon on Earth at angle A 1 second before Bob's photon arrives on Mars.

    What happens? Alice sees either an + or a - (1 or 0 or however you want to label it). There is nothing much for Alice to do here, as there is nothing different for Bob to see regardless of whether Alice does or does not measure her photon, and regardless of what angle A she choses to measure at.

    Bob sees a random pattern every time. + - + + - + - - - + - + - + - etc. Not much of a way to send a message. So you can see that your idea that Bob can tell the difference between a "polarized" photon and an
    "unpolarized" photon is incorrect. Because when Bob observes ANY photon, it will yield a polarization result.
  6. Nov 12, 2012 #5
    You are wrong about the 'reset' affecting both photons. It will only affect the photon you are doing it on, and will break entanglement.
  7. Nov 12, 2012 #6


    User Avatar
    Science Advisor
    Gold Member

    The reset mentioned by JohnLuck is just a colloquial way of describing things, and should not be taken too literally. The rule that actually applies is Malus.

    I would like to point out that the mechanism of breaking of entanglement is not fully understood. Therefore it is difficult to back up the comment that measuring one particle in the pair "causes" the break in the strictest of manners. Actually, measuring either one could be said to accomplish the same thing. That is because there is no sense in which time ordering of the measurements on the pair matters.
  8. Nov 12, 2012 #7
    It seems quite simple and obvious to me. "Entangled" means that the current polarizations along all possible angles are related (in this case, opposite). Measuring the polarization on one of the photons also changes it - at least along other angles. So measuring the polarization obviously breaks the relation, or breaks entanglement.
  9. Nov 13, 2012 #8


    User Avatar
    Science Advisor
    Gold Member

    So I would say what is simple and obvious is something of a matter of perspective. And your comment is a good first cut.

    Again, the devil is in the details here. For example, what is a measurement? If you use a beamsplitter to measure the polarization, but then recombine the H and V output properly, the original entangled state should be restored. If that is the case, then the original measurement did NOT cause collapse. See for example:

  10. Nov 14, 2012 #9
    On topic: If we change the entangled property, it is no longer entangled. That is the obvious part. Entanglement is a relation, not an influence. It does not allow us to transmit the change, or any information transfer.

    The fact that it may be possible to revert the change and restore entanglement does not matter. Whether we call the change a measurement or not does not matter. Phylosophical questions like "what is a measurement" and "does a tree make a sound" do not matter.
    At least not for the OP's porpose...

    But don't take that as a personal attack, DrChinese ;)
    The topic you bring up and the document you link is nevertheless an interesting read. Apologies to the OP for perhaps dragging the thread a bit off topic now, but I do have some comments about it that I will post here.

    Why do they bother to have pairs of photons and loops in both directions, instead of looking at one direction and just blocking the opposite beam in the corresponding loop on the right instead of on the left? The results should be the same... or should they not?
    Granted, it is an interesting idea to test out, but it significantly complicates the setup. They need perfectly separated photons and perfect detection and counting, etc... with a setup in just one direction they can easily make do with a regular beam and just measuring its brightness at the end.

    Why do they bother with the three different configurations, when only the first one is interesting? Seems that the whole ambiguity comes from the case where the first and third loops are blocked, while the second one is open. With Malus applied in a single step as if the middle loop didn't exist we get a different result than if we sum the results for the two possible paths when the loop exists.

    The real question is, does "recombining" succeed in restoring the previous polarizations (a), does it leave the photons in one of the split polarizations (b), or does it completely mess up and unpredictably affect polarization, making it appear random (c)?

    Recombining at the start of the chain is pointless as we begin with random, equal distribution of polarizations, and that is what we return to in every case.
    Recombining at the end of the chain is pointless as we never care what state the photons end up in when we count them.
    It is the recombining in the middle that is interesting.

    The simplified experiment that is interesting, with angles chosen for easy numbers then comes out to the following:
    filter at 0 deg, splitter and re-combiner at 30 deg, filter at 60 deg

    The resulting brightness factor should then be:
    (a): 0.5 * 0.25 = 0.125
    (b): 0.5 * 0.25 * 0.25 + 0.5 * 0.75 * 0.75 = 0.3125
    (c): 0.5 * 0.5 = 0.25

    What is the actual experimental result? That paper did not actually show it, and I apologize that I do not have the time to go search for other references...
    I assume that the re-combining actually can work like QM predicts, and we get the lowest number... otherwise we would not be having this discussion, as QM would have been ditched by now ;)
    It is curious if higher values can be received as well, from not-quite-perfect recombination effectiveness. And just how high do they get.

    But even with perfect experimental accuracy demonstrating a 0.125 factor, I still do not see the result as a clear demonstration of Bell inequality violation. More specifically, with the lack of clarity about how the recombination works and how Malus should be applied to the situation, I am not convinced that the derived inequality is supposed to be valid, even for a realistic and local model.
  11. Nov 14, 2012 #10


    User Avatar
    Science Advisor
    Gold Member

    There is an explanation, although you may or may not agree with the reasoning.

    Because of significant issues with detection and pairing of entangled photons, the "brightness" itself can be somewhat misleading. Instead, you could almost call it the "difference in brightness" that is being measured at particular angles. That should follow the QM prediction.

    Some of the issues with detection include the fact that often a photon arriving at one detector cannot be paired with a photon at the other anywhere close to the requisite coincidence time window. Now, is there some bias that would affect part of the universe and point us to a false result? The way around that is to also look for the output with the opposite polarization. This provides a convincing way to demonstrate that when one polarization stream decreases, the other is in fact increasing. Inference is not required because you can simply measure it.
  12. Nov 18, 2012 #11
    Ok, so someone suggested that are entangled and have opposite angles A, breaks their entanglement when we measure angle B and does not regain it after we measure angle A again.

    But this still allows for a machine as I see it. This time we still have a continuous stream of photons traveling to mars, only this time we use millions of photons pairs per packet. The photons that come out the right side of the machine we filter so that those who doesn't point up at angle A are destroyed and thereafter we send the photons into our mirror system. Those that come out of the left side of the machine we also filter at angle A but this time only the ones that do not point down are removed. This ensures that all the pairs we do not filter away are entangled in the same way. Then just an instant before the beam is detected on mars, we decide if we want to change the default configuration of the photon packet, which we can do by measuring angle B of the photon packet on earth. On mars they measure the angles of the photon packets and if they are all pointing down, we know the machine on earth did not change the default, but if they are all pointing in a more or less random direction, we know that they did. Thus we have send a bit of information. There will of course be a very low chance of either all photons being filtered away in a packet or having all photons pointing down, even though they were reset, but this is extremely unlikely and again just amounts to normal network noise, which we already have protocols to account for.

    I never said that changing the entangled property of one photon would change the other. The machine I describe above only needs to be able to "reset" an entangled photon at one angle and nothing else. Which paper did you read by the way, I got the explanation from here: http://www4.ncsu.edu/unity/lockers/users/f/felder/public/kenny/papers/bell.html
    (a bit patronizing, but written in layman terms)
  13. Nov 18, 2012 #12
    I'm not sure that I understand your idea. If you know what the state of the photon is, then it will not be entangled. You will just be sending a polarized beam of light to Mars and measuring a separate polarized beam on Earth.
  14. Nov 18, 2012 #13
    No, I know the state of the photon at angle A and it is entangled. I think you are confusing this with the uncertainty principle?

    Actually I don't know the angle of any of the actual photons, I only know that the photons with my desired angle will not have been intercepted and that it is very probable that more than 0 photons have gone through the filter.
    Last edited: Nov 18, 2012
  15. Nov 18, 2012 #14
    No it has nothing to do with the uncertainty principle. It has to do with the fact that preparing entangled particles is not very easy. If you need to measure the polarization in order to select which photons pass, you may no longer have entangled photons. If you extract the full information that can be encoded in the polarization after the entanglement, the photons will no longer be entangled unless there is some way that the information can be erased (there is a lot of evidence supporting this, but there is still some question if this is generally true and I don't know enough about all of the experiments ).

    Since I was having trouble understanding the particulars of your post, this may not be the reason that it won't work. If it makes sense to someone else, perhaps they can explain why it won't work.
    Last edited: Nov 18, 2012
  16. Nov 19, 2012 #15
    I do not know whether measuring polarization would destroy entanglement. Further, I do not know if filtering the photons counts as measuring. I will have to look at other experiments to find this out.
  17. Nov 19, 2012 #16
    I am repeating myself now, but you still seem to fail to grasp the obvious.

    A) After you pass a photon through a filter, you have broken entanglement, as the filter has changed the photon.
    (It may be possible to undo the change and restore entanglement if you merge the two possible paths of up/down photons just right, but that is irrelevant to your case.)

    B) Even if you did manage to filter and send a stream of photons with known A polarization to Mars that also have entangled counterparts on Earth, measuring the Earth photons on B does in no way affects the Mars photons. They will keep their known A polarization.

    How is it so hard?
  18. Nov 20, 2012 #17
    A) It is not obvious to me that filtering breaks entanglement. It depends on whether the filtering interacts at all with the photons that pass through it. Do you know the properties of all types of polarization filters?

    B) You are wrong. Measuring the angle B would set the angle A at random for all the photons. This is what is so special about entanglement, the entangled pair affect each other instantaneously over a distance, (Einstein called it "spooky action at a distance"). I suggest you read up on it. Also what did you think entanglement was?
  19. Nov 20, 2012 #18


    Staff: Mentor

    All polarization filters place the photons into an eigenstate, if they do not do that then they are not filters. Entanglement only happens when the state is a superposition of eigenstates, so it is self-contradictory to claim that you have placed a pair of photons in an entangled eigenstate.
  20. Nov 20, 2012 #19
    I... just...

    Well, good luck with breaking physics. I'm out.
  21. Nov 20, 2012 #20


    User Avatar
    Science Advisor
    Gold Member

    You are being unnecessarily rough with this comment.

    The fact is that your idea does not transfer any useful information faster than light. All outcomes appear completely random regardless of the angle you select to measure, and regardless of what you do elsewhere. No pattern emerges until and unless you compare Alice's result with Bob's. That requires a traditional information channel operating no faster than c.
Share this great discussion with others via Reddit, Google+, Twitter, or Facebook