[Full-disclosure] Interesting idea for a covert channel or I just didn't research enough?
mudge at uidzero.org
Thu Oct 6 15:22:07 BST 2005
What you describe would be a variant of 'dead-drop' covert channels.
Other examples would be:
. The use of public message boards where one program/person initiates
a connection out to the board and posts a message with particular
words/phrases/passages and another program/person scans said message
board looking for the 'encoded' messages.
. Web servers that have readable log files where one end of the
covert channel accesses certain web pages (existent or non-existent)
in a particular order and the other end looks for this message (or
alternatively has various combinations of request parameters in the
query strings which encodes commands or messages).
. etc. etc.
This type of covert channel has long been used by various governments
and organizations (think of clandestine messages being passed to or
from agents via personal ads). So, this is not a novel idea but you
are correct, there does not exist a tremendous amount of literature
on the subject - particularly in the public infosec/comsec communities.
A large percentage of 'dead-drop' covert channels will rely on a
shared 'code-book' between both parties.
On Oct 6, 2005, at 5:06 AM, PASTOR ADRIAN wrote:
> Sometime ago I thought of the following idea for a covert channel.
> Although the idea of covert channels is *not* new at all, I
> couldn't find anything in Google related to the following method of
> implementing a covert channel.
> The scenario is the following. The victim is a host with a host-
> level firewall which is blocking *all* incoming traffic. Somehow
> the attacker still needs to communicate with a backdoor planted in
> this host. Use a reverse shell and job done, you might say.
> Actually, there is another way which I thought would be more
> creative (IMHO).
> It works like this: the backdoor enables logging in the host-level
> firewall for all dropped packets, say Windows XP SP2 Firewall. Then
> the backdoor receives commands from the attacker by interpreting
> the properties of the dropped packets which were logged by the
> firewall. In other words, the backdoor is constantly reading the
> logs and parsing commands which were sent by the attacker embedded
> in packets which are being dropped (but logged) by the firewall.
> attacker sends packets -> packets are dropped by firewall ->
> packets properties are captured in logs -> backdoor reads logs and
> finds encoded commands -> commands are executed
> Now, for the way the backdoor would reply back to the victim is
> really up to you. One method that comes to my mind is by posting
> the responses to a PHP script which is located in some free-hosting
> webpage. The attacker would then access this webpage.
> Please, if you know anything related to backdoors intercepting
> commands from log files send me some links. Ideas, comments and
> flames are more than welcome :-) .
> pagvac (Adrian Pastor)
> Earth, SOLAR SYSTEM
> www.ikwt.com (In Knowledge We Trust)
> Full-Disclosure - We believe in it.
> Charter: http://lists.grok.org.uk/full-disclosure-charter.html
> Hosted and sponsored by Secunia - http://secunia.com/
Full-Disclosure is hosted and sponsored by Secunia.