1995-12-13 - Using quantum cryptography to effect covert channels

Header Data

From: Jas (Matthew K) <matt@lordmuck.itd.uts.edu.au>
To: cypherpunks@toad.com
Message Hash: 70d7f0626945fe481f553934525003706fc15885a1fa13d8de6a02fe4eae89d7
Message ID: <199512130321.OAA12208@lordmuck.itd.uts.edu.au>
Reply To: N/A
UTC Datetime: 1995-12-13 05:45:07 UTC
Raw Date: Wed, 13 Dec 1995 13:45:07 +0800

Raw message

From: Jas (Matthew K) <matt@lordmuck.itd.uts.edu.au>
Date: Wed, 13 Dec 1995 13:45:07 +0800
To: cypherpunks@toad.com
Subject: Using quantum cryptography to effect covert channels
Message-ID: <199512130321.OAA12208@lordmuck.itd.uts.edu.au>
MIME-Version: 1.0
Content-Type: text/plain


B2,3 and A1 system (Orange book), use covert channel analysis to
prevent leakages of information (intentional or otherwise). This forms
a part of enforcing MAC on such systems. well recently i thought of a
method to overcome these barriers so covert channels can again be used
to transmit and receive from such machines. this method involves using
quantum cryptography.

"randomization" is one method used to prevent covert channels, however
it is this "randomization" that can be used with quantum crypto to not
only make covert channels, but also makes the covert channels secure
as well as authenticated!

basically quantum crypto uses the fact that observing a system
modifies the system. this works on machines that use "randomization"
as a method of detering covert channels. ObExample, if an admin gets
suspicious of certain behaviour on a machine, the machine behaves
differently if he tries to investigate it (as a result of "poking
around" the machine to see what is going on).

this can be used with quatum crypto methods to covertly communicate
data (with security and authentification), and if someone attempts to
"observe" the behaviour on the said machines, it will break the
communication (because the machine behaviour changes, hence breaks the
message) rendering the message unreadable. this may slow down
communication some, and will need self-synchronization of some sort,
but at least you can communicate.

this idea is very rough however, and there may be gaping holes in my
arguments, but i think it might be an area someone may wish to look
into (especially the high assurance types). i really should write a
draft paper on this, but i dont really have a good enough
understanding of the formal methods to pull it off. however i have got
a coining name for this method, "covert quantum channels".

just a random thought to throw around...

			Matt
-- 
#!/bin/sh
echo '16i[q]sa[ln0=aln100%Pln100/snlbx]sbA0D3F204445524F42snlbxq'|dc;exit
Matthew Keenan   Systems Programmer   Information Technology Division
      University of Technology     Sydney Australia

It's nice to be in a position where people apologize because they
assume there's humor in your work, based on past experience,
but they're not sure where it is. -- Rob Pike





Thread