Zack Weinberg writes:
> I've seen claims that quantum key agreement lets both parties detect a
> man in the middle with no prior communication and no trusted third party.
Nope. The security of QKE relies on the parties both knowing a shared
secret key to authenticate messages. This begs the questions of
(1) how the parties communicated this secret---this doesn't have to
be a _prior_ secure channel but it does have to be a separate
secure channel;
(2) why the parties are bothering to use QKE to generate randomness
when they can much more cheaply generate local randomness and
send it through the separate secure channel; and
(3) why the parties are bothering to generate so much randomness in
the first place when they can much more cheaply use the key as an
AES key to encrypt and authenticate messages.
See http://cr.yp.to/talks/2009.10.06/slides2.pdf for a more detailed
cost-benefit analysis.
---D. J. Bernstein
Research Professor, Computer Science, University of Illinois at Chicago
_______________________________________________
cryptography mailing list
cryptography@randombit.net
http://lists.randombit.net/mailman/listinfo/cryptography
