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

Why can't quantum crypto be broken this way

  1. Feb 7, 2005 #1
    Quantum cryptography depends upon the notion that a sender transmits photons of known polarity, then the receiver choses random polarizations to filter the incoming transmission through. As each photon is either accepted, therefore telling the receiver that he correctly guess the sent polarity, or rejected, the receiver takes note. At the end of the first stage, the receiver tells the sender which photons he received and, with both of them knowing the polarity of those photons, they have successfully communicated n bits of information, encoded as the polarities...

    Here's my problem: the whole thing depends upon the assumption that a listener who attempts to determine the polarities of photons en route is destroying information - he has to send photons through a polarizer to determine their orientation. He cannot observe the communication and ensure that the intended receiver gets everything he was sent.

    What if the listener sends the transmission through a stimulated emmission medium? Stimulated emmission evidently gives you two exit photons - each with the same qualities as the input photon - direction and polarity included. As a listener, I could copy the photons in transit, giving me plenty of spares to determine the exact orientation of the origional, then send off a copy to the intended receiver. Voilla - I snoop on the backchannel where the receiver tells the sender what he received and I know exactly what their shared secret is.

    The only problem I see with this is if a pair of (stimulated) emmitted photons are either entangled or their polarity is not related to the stimulating photon's polarity, which is contrary to what I've been able to discern. Where am I (or the QC guys) wrong on this?

    This is right at the bleeding edge of my understanding of quantum. Please keep the replies at the appropriate level of pedantics. Thanks...

  2. jcsd
  3. Feb 7, 2005 #2
    Well, first of all, how would you know you have determined the exact polarization ?


    I wrote a text on this in my journal. I suggest you check it out.

    ps : it is not polarity but polarization...

    When eavesdropping, you have a 50/50 % chance to chose the right direction out of two possibilities (this is maximal chaos , right !).
  4. Feb 7, 2005 #3
    https://www.physicsforums.com/journal.php?s=&journalid=13790&action=view [Broken]

    Read the faster then light communication entry

    Last edited by a moderator: May 1, 2017
  5. Feb 7, 2005 #4
    If a QC communication uses 4 different porarization directions (0, pi/4, pi/2, 3pi/4) and you make 4 copies of the photon, you can check each of these. Whichever one passes tells you what the original was.

    Doh. Just got too much into the details and blew the english....

    Usually, 4 directions are used - it makes it easier to detect if you have an eavesdropper, among other things.

    I'll look at the doc you linked in your second post and see if it's helpful to me.... thanks...
  6. Feb 7, 2005 #5
    That would be the Mozart Requiem. God I'm such a geek - not only do I recognize the blasted thing but I can (very roughly) translate it.

    *sigh* back to work.
  7. Feb 7, 2005 #6
    No, there are indeed 4 possible polarizations. However, when measuring such polarizations you need to make a choice. Either you measure the horizontal and vertical polarizations or either you measure the diagonal polarizations. However, once you made your choice (50/50 % chance of being correct) you can either be right or wrong. I mean, the photon that you measure will be polarized according to ONE of the TWO possible sets of polarization. You cannot measure both sets at the same time. If you are wrong, you disturbe the message because the polarization will have jumped from one direction of set 1 to another direction of set 2. Eg : from vertiacl to diagonal...


    ps : it is not only the Mozart reqium although this text is also sung there. It is just an extract of the reqium-mass which has a fixed content and structure. Though i must say that some composers have changed the words. You ain't no geek, you just have lot of culture :approve:
  8. Feb 7, 2005 #7
    Okay, it sounds like we're moving in the right direction. It is true, then, that one can make however many identical copies of the photon they like, but something between that action and passing them through different polarizing filters is my problem? I suspect that separating them into different streams for testing would cause an issue....
  9. Feb 7, 2005 #8

    Cloning ???

    You cannot make many copies of one atom. This is forbidden by QM because the actual copying itself would alter the state of the atom that you are copying. So you can forget about making many copies. Here is your biggest problem.


    In other words : when trying to copy, you can ensure yourself from the fact that the copied atoms are exact duplicates of one and other
  10. Feb 7, 2005 #9
    AGH! No! Stimulated emission of photons - as in the effect used to generate a coherent laser beam.
  11. Feb 7, 2005 #10
    Indeed, in stimulated emission the photons are exact copies of one and other. The first photon triggers the emission of a second photon with the same properties as the first photon. However this laserbeam is not the same as a bunch of photons of which you can "take one photon at the time"


    And remember you choice of polarization is also a very big problem. Would you trust on a "guessing-mechanism" that has the same reliability as flipping a coin four times in a rowe and guessing the outcome ???
  12. Feb 7, 2005 #11


    User Avatar
    Staff Emeritus
    Science Advisor

    Actually, I don't believe it is true that you can clone a photon with 100% accuracy. Unfortunately, I don't recall the exact details but IIRC the stimulated emission process has some probability of producing a clone, and another smaller probability of not producing a clone.

    I found some links on the topic
    http://www.duke.edu/~mag18/cloning/Resource%20Links.htm [Broken]
    Last edited by a moderator: May 1, 2017
  13. Feb 8, 2005 #12
    That is correct

Share this great discussion with others via Reddit, Google+, Twitter, or Facebook