Re: Reply-block based pseudonym systems are broken? [Was: Re: SURB-Format]

[tainaron@xxxxxxxxxxxxxxxx]

Hi,

jeffery wrote:
> How trivial is this type of attack?

Everybody trivially controls the nym end of the chain
(just send a message to the nym account), and it's sufficient if
you can just count messages delivered to the regular account of
the suspected nym holder, e.g. through the regular mail server
log... Perform the attack, seize the log file and you're done.

This provides no starting point to find out the nym holder, but
it is effective as soon as you have narrowed down your suspicions
to a few individuals.
This attack does not depend on the volume of traffic sent over
the network in general, but on the volume of traffic received by
the nym. It might help to generate cover traffic, but that's
wasting bandwidth, too.

Sufficient messages are between 37 in the best and 1775 in the
worst case (as stated in the paper), and assuming identically
behaving users this would usually be an overestimate.

The nym holder might or might not figure it out, the attack
could easily be disguised as spam surge. Or he might not find
out in time. And even if he does, what to do about it?

Ciao

Tobias
-- 
mbox:admin@tainaron