Cryptography Reference
In-Depth Information
long. Researchers at Los Alamos achieved 48 km in 1999. This is an outstanding
achievement in view of the fact that
single
photons are transmitted.
Unfortunately, even such outstanding achievements don't remove all problems.
If Alice and Bob are at a distance of 2000 km, then quantum cryptography
won't work over fiberglass even if the most sophisticated technology were
used. The solution could be airborne, at least partly: through the atmosphere
to a satellite and back to earth again.
Is this doable? Isn't air an optically 'dirty' medium compared with superclean
fiberglass? Richard Hughes of the Quantum Information Team at Los Alamos
has been working at it for years and has achieved results that stun physicists
with disbelief. A receiver with a diameter of a few centimeters, circulating the
earth in 8 km/s at a height of 300 km, would have to be hit by one single photon
that must neither be absorbed nor change its polarization as it travels through
air. Hughes hasn't gotten there yet, but almost. He uses the following tricks:
•
The selected wavelength is 770 nm (this is borderline infrared). The
absorption is very low in this range, and turbulences (that could change
the polarization) have a smaller impact due to their typical expansion of
several centimeters.
•
Light scatter is widely excluded by receiving only photons that ideally
come exactly from the sender's direction. To this end, a special receiver
unit was built at Los Alamos.
•
A sharper frequency filter lets pass only photons of the transmitted wave-
length.
•
Nevertheless, many photons are still disturbed, though they do come
from the right direction and have the right frequency. For this reason,
a time window of only 5 nanoseconds is opened for reception in every
microsecond.
•
Other sources of interference are air turbulences. Though they don't
absorb the photons, and they don't change their polarization noticeably,
they provoke a change to the photons' traveling time due to density fluc-
tuations in the air. However, this is not jerky. The time difference remains
moderate within 1 microsecond.
To compensate for such differences in the time of flight, each 'sharper'
pulse is preceded by a 100-nanosecond measuring pulse. This is how the
current time of flight is determined, and the receiving window is kept
open at exactly the right point in time.
