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

Texas A&M Photonless Communication

  1. Apr 22, 2013 #1
  2. jcsd
  3. Apr 22, 2013 #2


    User Avatar
    Science Advisor
    Gold Member

    Here is the full article:


    "It is not difficult to see that if Bob blocks the photon, detector D 2 clicks. Also, if Bob passes the photon, detector D 1 clicks. ...

    "In summary, we strongly challenge the longstanding assumption that information transfer requires physical particles to travel between sender and receiver by proposing a direct quantum communication protocol whereby, in the ideal asymptotic limit, no photons pass through the transmission channel, thus achieving complete counterfactuality. "

    As I read it, they propose an experimental setup to allow such communication. On the other hand, I question whether the implementation will work quite as they describe. If it did, it seems to violate the no-signaling rule. So this is going to be a controversial result.
  4. Apr 22, 2013 #3
    Aren't there other situations in which the signalling rule has been violated? Both local realism and CFD seem to be at the heart of entanglement. Are there other situations?

    With that said is this simply just another reason local hidden variable ideas should be abandoned and our notions of information exchange rethought?
  5. Apr 23, 2013 #4
    Figure 2a from the paper

    I read the experimental setup of figure 2a in the paper, but I must be missing the significance of this. Let me explain.

    When Bob chooses to turn SW off, then as you reach the bottom of the diagram 2a most of the amplitude of the wave function is passing through SW which presumably is in the vicinity of Bob (under his control). Eventually this wavefunction reaches D2, which presumably is in the vacinity of Alice, and clicks. Hence most of the wave function passes from Alice to Bob and back to Alice (actually several times) to communicate the "logic 0". Given this, I don't understand why this form of communication called counterfactual. Is there is a rigorous definition of what counterfactual means in this context?

    Is this any more counterfactual than a much simpler form of communication where there's a light source behind Bob with a path to Alice which is blocked most of the time, but when Bob wants to send "logic 0" he just opens the path? This too requires the exchange of a wave function to communicate "logic 0" (but not "logic 1") and Bob never needs to "collapse" the wave function or measure anything to communicate.
  6. Apr 24, 2013 #5
    wheelersbit, in my opinion it's much easier to understand this by starting with Figure 2a and the description of 2a which starts on page 2. Those beam splitters (BS) are setup to reflect light most of the time (keep the photons on the left), but the reflectivity is tuned just right so that, by the time a photon reaches the detectors Di at the bottom, all the wave function is actually on the right side and enters D2 (assuming all SWs are "off"). However, if Bob turns the SWs on, then the wave function is "collapsed" after each BS and, due to the very high reflectivity, almost always collapses on the "left side" and the math works out that, even after N BSs the wave function is almost entirely on the left side (when the SWs are "on"). Hence, Bob can (almost perfectly) control which detector clicks by turning the SWs on or off.
    Last edited: Apr 24, 2013
  7. Apr 26, 2013 #6
    There's been some updates regarding this paper. Firstly, it has finally been published (after almost 1 year since it appeared at arXiv) in PRL. Secondly, there is a comment on this paper by Lev Vaidman (see http://arxiv.org/abs/1304.6689) where he argues against their claim to some extent. I haven't had time to read it though so I can't comment on the validity but you might be interested in checking that out.
  8. Apr 26, 2013 #7
    Thanks Zarqon! It looks like Vaidman is making the same argument -- the "logic 0" communication requires most or all of the wave function amplitude to pass to Bob.

    I still feel like I am still missing something because the argument seems too obvious, yet the paper was reviewed and published. Anyone have a deeper understanding of why this would have been considered counterfactual communication?
  9. May 1, 2013 #8


    User Avatar
    Science Advisor

    Nicolas Gisin also uploaded a comment to the ArXiv: http://arxiv.org/abs/1304.8053.

    In a nutshell his conclusion is that the result is not as spectacular as it sounds.
  10. May 2, 2013 #9
    Simply put..... as an illustration:

    If you put a broker (Charlie) in-between Alice (seller) and Bob (buyer). Bob and Alice don't need to communicate to complete the property transaction.

    Essentially Alice is sending the information to Charlie who sends it to Bob.

    Alice and Bob, obviously, don't need to communicate directly. Charlie is doing the job. Although Charlie and Bob have a complicated/sophisticated way/process/heuristic to communicate.

    or take an example of a translator

    Alice gives a document in French/Chinese to translator (Charlie). Charlie translates it in English and gives to Bob.

    Alice and Bob did not need to communicate directly.
    Last edited: May 2, 2013
  11. May 2, 2013 #10


    User Avatar
    Gold Member

    http://arxiv.org/abs/1304.7469 "Asking photons where they have been."
    Here is another paper by Vaidman and colleagues in which they argue that we must consider backward-in-time prpogation of wavefunctions to understand this type of experiment.
Share this great discussion with others via Reddit, Google+, Twitter, or Facebook