|
|
Quantum Information Research at NIST: Goals and Vision |
Quantum cryptography allows two parties, known as Alice and Bob, to exchange information in a manner that is, in principle, secure. By using single photons to encode information, Alice and Bob can detect attempts at eavesdropping. NIST is demonstrating prototype technologies for practical distribution of secret quantum keys, which are used to encrypt and decrypt messages. The NIST system sends and receives photons in four different orientations to represent the values 1 and 0. Each photon is sent in one of two modes, either vertical/horizontal orientations of the electric field, or plus 45 degrees/minus 45 degrees orientations. Within each mode, one orientation represents 0, and the other represents 1. To visualize how this works, imagine that each photon is a tiny envelope moving perpendicular to the ground (vertical=1), parallel to the ground (horizontal=0), tilted at 45 degrees to the right (plus 45 degrees =1) or tilted 45 degrees to the left (minus 45 degrees=0). Each photon fits best through one of two types of detectors, or “mailboxes.” Alice randomly chooses both a mode and an orientation for each photon. Bob randomly chooses between the two modes when he tries to detect a photon. This can be visualized as choosing a mailbox slot that accepts only envelopes flying in certain orientations. If he chooses the same mode that Alice used for a particular photon, then Bob always measures the correct orientation, and hence, its bit value. But if he chooses a different mailbox, then he may get the wrong bit value for that photon. To make a shared “key” from a stream of photons, Alice uses a conventional communications channel to tell Bob which mode she used for each photon (without revealing its bit value). Bob tells Alice which photons he measured using the correct mailbox (but again, not sharing their values). Then they both discard the other bits, the ones Bob measured with the wrong mailbox. The correct measurements constitute the secret key that Alice and Bob now share. If someone, generally referred to as Eve, tries to eavesdrop on the transmission of the stream of photons, she will not be able to “read” it without altering it. When she measures a photon, it is converted to electrical energy and destroyed. This photon will not reach Bob and will not contribute to the key. Eve may send a replacement photon, but, because she may have used the wrong mailbox to intercept the original photon, she may be wrong about the bit value. Thus, for photons detected with the wrong mailbox mode, Eve may introduce errors into the key constructed by Alice and Bob. When Alice and Bob detect an unusual number of errors they will be alerted to Eve’s presence. The illustration shows an example. If Alice sends a photon in the vertical orientation, and Bob chooses the correct mode or mailbox (the + mode), then he will always measure the photon correctly and get the correct bit value of 1. (See top of illustration.) On the other hand, if he chooses the wrong mailbox (the X mode), then he might get the wrong bit value when he measures the photon. In this case, Alice and Bob will find out later that they used different modes, and discard that bit value in making their shared key. If Eve intercepts
and resends the photons, errors occur in the data that alert Alice
and Bob to the eavesdropping. In the inset picture, Eve tries to receive
a vertical photon in the wrong mode and measures it incorrectly as
a -45 degree photon. Eve then sends a -45 degree photon to Bob. Bob
receives the photon in the vertical/horizontal mode and records a horizontal
photon. He used the correct mode but obtained the wrong measurement
result and the wrong bit value. When Alice and Bob compare their sending
and receiving modes they will end up with errors in the key (red data
in the scorecards in the illustration). A NIST error-correction method
will detect these errors, and Alice and Bob will know that Eve has
been listening. |
||
|
Date
created: 4/16/06
|
