NODE 0b0af3f5Re: Introduction- Telephone
"Ron Bardarson" <Ron_Bardarson@smtp.esl.com>Mon, 12 Sep 94 08:25:32 PDT
Reply to: RE>>Introduction: Telephone traffic analysis
>> I stumbled upon this great explanation of telephone traffic analysis and
>> thought I'd share it with the list:
>It is to evade this sort of traffic analysis that complex routing, batching
>and clustering of anon remailer traffic is being evolved.
As long as 'someone else' controls the medium thru which the messages pass,
traffic analysis will be possible. It may be a large problem, but it will
still be doable with sufficient motivation. When you have an untappable wire
between the sender and recipient, then traffic analysis is impossible (and
crypto unnecessary). In order to get to the remailer chain you mention, I
(for instance) can ONLY pass thru a couple a machines, which if properly
monitored, reveal the traffic flow despite the remailers. You can still beat
traffic analysis (and you still need crypto), but not in this medium.
NODE 1a098ca5Introduction- Telephone
hughes@ah.com (Eric Hughes)Mon, 12 Sep 94 09:28:16 PDT
When you have an untappable wire
between the sender and recipient, then traffic analysis is impossible (and
crypto unnecessary).
Yet you can create an untappable wire with cryptography! The device
in question is called a link encryptor. Take a stream cipher and run
it continuously across the channel in question. Pad the asynchronous
traffic when it's not flowing and add some synchronization to both the
stream and the data insertion.
You can tap the physical line still, but the interceptions reveal zero
information (computationally--the stream cipher _is_ keyed, after
all).
A good project would be virtual link encryptors for the Internet.
Eric