1997-10-01 - Traffic Analysis and Cover Traffic

Header Data

From: Tim May <tcmay@got.net>
To: “Robert A. Costner” <cypherpunks@cyberpass.net
Message Hash: ea65ccc97d522c6b49693b96dad4320a039c966f25de933a90ddc54b57ddc3b1
Message ID: <v03102803b05829bdf10b@[207.167.93.63]>
Reply To: <199710010759.JAA27558@basement.replay.com>
UTC Datetime: 1997-10-01 16:54:33 UTC
Raw Date: Thu, 2 Oct 1997 00:54:33 +0800

Raw message

From: Tim May <tcmay@got.net>
Date: Thu, 2 Oct 1997 00:54:33 +0800
To: "Robert A. Costner" <cypherpunks@cyberpass.net
Subject: Traffic Analysis and Cover Traffic
In-Reply-To: <199710010759.JAA27558@basement.replay.com>
Message-ID: <v03102803b05829bdf10b@[207.167.93.63]>
MIME-Version: 1.0
Content-Type: text/plain



At 2:03 AM -0700 10/1/97, Robert A. Costner wrote:

>Let me get this straight.  You are suggesting that anyone who wishes to be
>anonymous should send a continuous 24 hour stream of low bandwidth data to
>a central point in an effort to help keep anyone from knowing that they
>wish to be anonymous.

He was describing how a constant-traffic pipe defeats "sudden burst of
activity" types of traffic analysis. (As, for example, when nightime
activity in the White House is signalled by deliveries of lots of pizzas.)

In fact, this constant traffic approach is the basis of PipeNet, proposed
and described by Wei Dai, and, I think, being implemented.

Obviously not everyone will want this, or want to pay for it, etc. Nothing
surprising there. But some may. In particular, it makes more economic
sense for sites already full-time on the Net, e.g., remailers not on
dial-up lines. (And even on dial-up lines, a call every hour, with a packet
of traffic, can implement this cover traffic scheme, albeit in a different
form.(

>
>While this may help correct the latency problem, how do you think this will
>effect anonymity?  Do you think that by sending a continuos stream of data
>to the remailer, the sender will be less identifiable?

Again, _latency_ per se is not important, _mixing_ is.

As I understand your example above, the sender becomes less identifiable
because he can always be counted on to send some packets; an attacker
cannot see a message after a long period of no messages and correlate it to
a similar sudden increase of activity at a possible recipient machine.

Traffic analysis is something you should look at. And think about.

--Tim May


The Feds have shown their hand: they want a ban on domestic cryptography
---------:---------:---------:---------:---------:---------:---------:----
Timothy C. May              | Crypto Anarchy: encryption, digital money,
ComSec 3DES:   408-728-0152 | anonymous networks, digital pseudonyms, zero
W.A.S.T.E.: Corralitos, CA  | knowledge, reputations, information markets,
Higher Power: 2^2,976,221   | black markets, collapse of governments.
"National borders aren't even speed bumps on the information superhighway."








Thread