Nmap Development mailing list archives

RE: Nmap OS Database and Artificial Neural Networks


From: "Dario Ciccarone (dciccaro)" <dciccaro () cisco com>
Date: Mon, 18 Feb 2008 22:30:17 -0500

Interesting. But - the IEEE expects for people to pay for your paper. While I don't specially mind, I think you're 
leaving part of the audience out that way.

Nice graph - now, how about also making a copy of the paper available at ufrn.br too ?

Dario
 

-----Original Message-----
From: nmap-dev-bounces () insecure org 
[mailto:nmap-dev-bounces () insecure org] On Behalf Of João Medeiros
Sent: Monday, February 18, 2008 8:45 PM
To: Nmap Dev
Subject: Nmap OS Database and Artificial Neural Networks

Hi all,

    Until April of last month I've developed a tool that convert an
Nmap OS Fingerprint database entry to a vector of normalized real
numbers. As result of this we construct a tool that has the following
features:

    - A contextual map that show the similarities among different
operating systems [0].
    - A new OS fingerprint metric that can be used with 
mathematical tools.

    Initially this work have focus on classification of unknown
operating systems of embeded devices like printers and something
others that has a TCP/IP stack implementation. Also with the map and
the use of Euclidean distance in the converted fingerprint database we
can confirme that some softwares share a similar TCP/IP stack
implementation (like Mac OS X and FreeBSD). This kind of operating
systems contextual map was produced for the first time in this work.
Because this I'm telling all you. If you have any ideia with the use
of this map we can tell us. The initial published paper about this
work can be view at IEEE search site [1].

    Not only the map, but the convertion of Nmap fingerprint database
into real vectors can grow the use of mathematical tools to help os
fingerprint task. Using pattern classifiers like MLP (Multilayer
Perceptron), RBF (Radial Base Functions) or SVM (Support Vector
Machines). Just the use of Euclidean distance showed its potential.

    Resumimg, due to sponsors constraints I can't release the code
now, but it helpful if you show to us your ideas. Thanks for your
time.

Att, João Medeiros.

[0] - 
http://www.dca.ufrn.br/~joaomedeiros/tmp/2008-01-30_21-05-08_1
0000-20x20.html
[1] - 
http://www.ieeexplore.ieee.org/xpl/freeabs_all.jsp?isnumber=44
16743&arnumber=4416854&count=223&index=110

_______________________________________________
Sent through the nmap-dev mailing list
http://cgi.insecure.org/mailman/listinfo/nmap-dev
Archived at http://SecLists.Org


_______________________________________________
Sent through the nmap-dev mailing list
http://cgi.insecure.org/mailman/listinfo/nmap-dev
Archived at http://SecLists.Org


Current thread: