RE: Proving non-repudiation in e-Commerce App

["Craig Wright" <cwright () bdosyd com au>]

Hello,
Firstly there is no way to prove non-repudiation. There is no valid means to prove encryption. For those who do not 
agree, please read up on computational mathematics and the N vs NP problem (also see computational theory in general). 
Micheal Sipser (from MIT) has some excellent papers on the topic.

Next lets get to prove. Prove is a mathematical determination of a rule. Even in the case of a discovered and somehow 
proven encryption algorithm there is no way to prove non-repudiation.

What does this mean? It comes down to a likelihood determination. This is a probabilistic determination of the 
Cumulative distribution function (CDF) associated with the survival and hazard functions of the plot of time against 
likelihood of compromise.

Even in cases of a perfect algorithm there is an associated hazard function associated with a brute force compromise of 
the key. In most cases this Probability density Function (PDF) correlates to a Poisson distribution.

So what you are looking at in reality is a survival function that will be acceptable in a court of law that will not be 
readily repudiable by the opposing party.

To do this you need to look at proof beyond reasonable doubt. This is due to the criminal standard of proof being used 
for deceit. As you wish to prove against a person who may be lying this is the necessary level of proof. In common law 
courts this is generally (though not exclusively) held at a determined confidence level (CI) of 99%. That is an alpha 
set at 1%.

Now the determination needs to be complete in a cumulative manner which includes the totality of the systems. In this 
you need to determine the individual hazard function for each of the components. This is than extrapolated into the 
total Survival function estimate for the system.

One property of the exponential distribution and hence the Poisson process is that it is memory-less (This is the 
number of incidents occurring in any bounded interval of time after time t is independent of the number of arrivals 
occurring before time t).

Now this means that you are attempting to determine the lambda function λ(t) associated with each hazard occurrence 
(being the likelihood of brute force or other key compromise). The number of expected key compromises for each 
component is than the integral of λ(t) for the period from 0 (start time) to a determined safe time (i.e. promised 
non-repudiation of 5 years, 25 years etc). 

So yes there are ways to achieve what you are asking. What you are looking at is the expected "safe" time of your 
system.

Regards,
Craig

-----Original Message-----
From: Joe [mailto:bitshield () gmail com]
Sent: Friday, 2 June 2006 4:32 AM
To: security-basics () securityfocus com
Subject: Proving non-repudiation in e-Commerce App

Dear List-Members

I'm currently dealing with a review of an e-Commerce Application. One
goal is to prove that this application properly implements a
non-repudiation mechanism throughout the whole process-flow. This flow
starts at the user authentication, communication over the web to the
server component, then processing of the client requests and finally
logging.

The non-repudiation has similarities with e-Banking which points me to
the following keywords: digital signature, signed logging and time
stamp protocols. Using Google I also found various sources discussing
most of those points individually. However I'm looking for a more
general, broad and complete approach.

Do you guys have interesting sources and experiences about verifying
non-repudiation? Are there standards, defined processes, work-flows,
and implementation- or audit guidelines?

Thanks for your feedback
Joe


Liability limited by a scheme approved under Professional Standards Legislation in respect of matters arising within 
those States and Territories of Australia where such legislation exists.

DISCLAIMER
The information contained in this email and any attachments is confidential. If you are not the intended recipient, you 
must not use or disclose the information. If you have received this email in error, please inform us promptly by reply 
email or by telephoning +61 2 9286 5555. Please delete the email and destroy any printed copy. 

Any views expressed in this message are those of the individual sender. You may not rely on this message as advice 
unless it has been electronically signed by a Partner of BDO or it is subsequently confirmed by letter or fax signed by 
a Partner of BDO.

BDO accepts no liability for any damage caused by this email or its attachments due to viruses, interference, 
interception, corruption or unauthorised access.