a new way to do anonymity

Johnathan Corgan jcorgan at aeinet.com
Tue Feb 7 19:29:22 PST 1995


>This is what I wrote:

>> Imagine a server that allows you to open a 
>> low bandwidth (let's say around 100 cps, in order to reduce costs)
>> link-encrypted telnet session with it, and provides you with a number 
>> of services, for example a link-encrypted talk session with another 
>> user.  You'll need to maintain the link 24 hours a day to defend 
>> against statistical analysis, and of course you can chain a number of 
>> these servers together in a way similiar to chaining remailers.

There are many, many analogies you can draw about a network of this type
to an ATM (asynchronous transfer mode) network.  By simplifying just slightly
from what you describe to only include an encrypted, switched-pipe methodology,
you now have a "cloud" type network with service entry points that are defined
by a pair of byte streams (one in each direction).

The switched path could be set up and torn down dynamically by the user by
interacting with the "switch" at each point to select the next hop the 
encrypted byte stream will follow.  Of course, just like in remailer chaining,
the data that indicates which hop to follow is encrypted with the data in a
form only the switch can decrypt.  Alternatively, once a path is set up between
switches, it can be assigned a virtual path identifier that has only local
significance at each hop, with the switch performing a lookup to forward
packets and substituting a new path number with significance at the next hop.

The above description is pretty unclear, I think, but many of these concepts 
have been fleshed out to a significant amount of detail in ATM circles.

Fixed length data packets (at the encrypted telnet level) also make it very
easy to aggregate individual circuits into higher bandwidth pipes that
connect server to server.  With these continuously running with cover traffic,
individual circuit establishment is much more immune from traffic analysis.
Cover traffic is substituted with real traffic as necessary, up to the bandwidth
of the pipe.

To summarize, what has been described is a method to establish a "network within
a network", using encrypted telnet, to provide a connection oriented, unreliable
packet switched link layer protocol.  Sounds remarkably similar to IP (except
for the connection oriented portion of it.)

What can you do with a network like this?  By layering a TCP process on top of 
this "Pipe-Net" IP like service, any of the standard TCP based application 
protocols can function between two end point systems, such as SMTP, FTP, HTTP,

What is so neat about this is that it could probably be done in user space,
and since the packet based protocol is defined as unreliable, switches could
come and go, with some sort of switch-to-switch protocol that propagates route

Eric, you could probably chew on the trust implications of all this.

Perry, I'm sure all the IPSP/SSL/SOCKS/whatever stuff you know so well could
provide a lot of building blocks for this type of thing.

Wei, your traffic analysis treatment of this sort of thing would go a long way
toward uncovering weaknesses and determining operational requirements and 

Tim, what massive social effects would it have if this type of network service
were to become widely deployed? :)

At first glance, this Pipe-Net idea doesn't seem to take a lot of rocket science; 
it seems that most of the components or algorithms are are already in use, just 
in a very different way.

I can think of a number of problems already, however.  Spamming.  Bandwidth
limitations.  Complexity of client and switch software.  Standards.  Flow

In other works, all the stuff the ATM forum is already dealing with :)

Come to think of it, has anyone thought of something like this before?

Johnathan Corgan       "Violence is the last refuge of the incompetent."
jcorgan at aeinet.com                    -Isaac Asimov
WWW:                     ftp://ftp.netcom.com/pub/jc/jcorgan/home.html

Version: 2.6.2


More information about the cypherpunks-legacy mailing list