Re: Quantum Cryptography can not work

I was NOT refering to imperfect implementations of QC, that fail
to use true (enough) single-photon sources, which I am told
has existed in would-be commercial systems.

I was refering (without any real understanding) to something
called Breidbart Basis or Breidbart Eavesdropping that, I was
told, could break BB84 as first stated, or is it variants
of BB84 that published articles have been considering.

When we deal with perfect devices, the situation with BB84 is
pretty simple as we are in the ideal case: perfect alignement,
single source pulses, no noise in the line etc. With no noise in the
line if Eve do an "intercept-and -resend" attack (by measuring
a photon) whatever basis she chooses (even the Breidbart ones), she
introduces some noise in the line and thus reveals her presence.

However in real world we don't have single source pulses (a small
fraction of pulses will have more than one photon), the optical
alignement is not perfect and the detectors are noisy. So we use
special algorithms (error correction and privacy amplification) to
correct error and reduce Eve's information. The difficulty is to set
an upper bound of Eve information to set privacy amplification
parameters. For that during the last decade there was a lot of
research to analyze what Eve's best strategy is in case of imperfect
devices and what amount of information she can get.

ouah


.

Quantcast