Cryptography Reference
In-Depth Information
5.9 Experimental Postselection
In our demonstration experiment, the setups of Alice (Figure 5.8) and Bob
(Figure 5.10) are separated by roughly 30 cm. To show the independence of
two measurements in the case of a beam splitting attack, the signal is di-
vided at a 50 : 50 splitter and measured by two Stokes detector setups (see
Figure 5.13). This independence is crucial for the postselection procedure as
described in Section 5.2. When both detectors are set to measure in the same
basis (in this case
S
2
), an unmodulated
S
1
polarized coherent state gives a
Gaussian distribution of the measurement results in each detector. The re-
sults are uncorrelated, as can be seen from the circular correlation plot in
Figure 5.13 (left). For an
S
2
modulation with low amplitude, the plot (Fig-
ure 5.13 right) shows a slight ellipticity, revealing small correlations between
the detected signals. Still it is not possible to deduce the sign of the mea-
surement result of one detector from the outcome of the other detector with
certainty. A potential eavesdropper Eve who uses detector 2 cannot infer the
results of detector 1, even though she has measured 50% of the signal. Note
that the setup of Figure 5.12 was also used with one
S
2
and one
S
3
detector to
produce Figure 5.4.
In the rest of the experiments, only a single detector with electro-optical
basis switching was used (cf. Figures 5.11 and 5.12). The losses due to nonunity
photodiode efficiency, EOM transmissivity, and optical imperfections in the
detector were treated as if they were transmission channel losses. This con-
servative point of view implies that Eve could manipulate Bob's receiver and
increase its efficiency while sending more imperfect states at the same time,
gaining additional information. With an additional attenuator in between
Alice and Bob, overall transmission levels of 79% and 36% were set. The
modulation was adjusted to give an average coherent amplitude of 0.6 in the
0.4
0.3
0.2
0.1
0.0
-0.1
-0.2
-0.3
-0.4
-0.4
0.4
0.3
0.2
0.1
0.0
-0.1
-0.2
-0.3
-0.4
-0.4
-0.3
-0.2
-0.1
0.0
0.1
0.2
0.3
0.4
-0.3
-0.2
-0.1
0.0
0.1
0.2
0.3
0.4
S2 (detector 1)
S2 (detector 1)
Figure 5.13
Signal divided by a 50:50 beam splitter and measured by two
S
2
detectors.
Left: no
S
2
modulation of the signal. Right: small
S
2
modulation, correlations of the
two detected signals result in elliptical shape of the scatter plot. As detector 1 and
detector 2 have slightly different gains, a small horizontal ellipticity is introduced in
both graphs.
