Cryptography Reference
In-Depth Information
Using entangled states, such a correlation would prevent from violating a Bell
inequality and could thus be discovered easily.
Although all Bell experiments intrinsically contain the possibility for
entanglement-based QKD, only a few experiments have been devised in order
to allow a fast change of measurement bases. Interestingly enough, the first
experiment that fulfills this criterion is the test of Bell inequalities using time-
varying analyzers, performed by Aspect et al. in 1982 [55] — at a time when
quantum cryptography was still unknown, even the single-photon based ver-
sion. More experiments enabling active [56] and passive [41] change of bases
followed in 1998 and 1999. However, as with the first-mentioned experiment,
the bases chosen for the measurements are chosen in order to allow a test of
Bell inequalities and not to establish a secret key. The first experiments that
allowed the distribution of a quantum key were finally performed in 2000
[57-59], and more followed in 2001 [60], 2002 [61] and 2004 [43,44].
All our experiments [43,44,59,60] incorporated a passive choice of bases.
In 2000 and again in 2003, we took advantage of an original solution offered
by time-bin entanglement [44,59]. As mentioned before, only detections in the
central time window correspond to a projection on a basis with eigenstates
represented on the equator on the qubit sphere. Interestingly, detections in
one of the two remaining windows are not unwanted events that have to
be discarded after postselection but correspond to projection onto the
|
0
or
states. Hence, time-bin entanglement offers passive choice of bases
for free. While the first experiment [59] has been performed over a short
distance, we could recently extend this distance to 50 km (fiber on a spool)
with actively stabilized interferometers, which is the longest transition span
for entanglement-based QKD to date [44].
In 2001, we realized a QKD system based on energy-time entanglement
[60], similar to the test of Bell inequalities mentioned in Section 2.4.2.1. How-
ever, this realization took advantage of an asymmetric setup, with the source
close to Alice's, instead of a setup designed for tests of Bell inequalities with
the source located roughly in the middle between Alice and Bob. This makes
it possible to employ high efficiency and low noise silicon avalanche pho-
todiodes detectors (that cannot detect photons at telecommunication wave-
lengths) at Alice's side, together with photons at nontelecommunication wave-
length (that cannot be sent through long fibers due to enhanced absorption).
The second photon (again at 1550 nm telecommunication wavelength) is
transmitted through 8.5 km of dispersion shifted fiber to a fiber optical in-
terferometer, equipped, as usual, with Faraday mirrors. The passive choice
of bases was implemented using polarization multiplexing (see Figure 2.10),
similar to Figure 2.7d. We recently repeated this experiment using 30 km of
standard fiber, managing the dispersion by filtering or compensation [43].
Finally, in addition to the mentioned two-party QKD schemes, we re-
ported in 2001 a proof-of-principle demonstration of quantum secret sharing
(three-party quantum cryptography) in a laboratory experiment [63]; see also
Chapter 7 of this topic. This rather new protocol enables Alice to send key
|
1
