Russell Shaw has located an analysis by Art Reisman, CTO of APconnections, maker of packet shapers. Apparently, Reisman was asked to see whether an external party can detect and if so then block Skype traffic. The background discussion is long, but the real analysis is short. It seems none of the tricks the author uses to identify a voice flow worked. So, the author surrenders and hopes others will solve the mystery. It is so disappointing. The author could have at least commented, if not studied, the scheme used by Verso. If you recall they have claimed that they can successfully block Skype traffic, that some service provider in China is trialing their product and the product has been verified by an independent agency. Either the author is correct and Verso’s claim is not valid or the author’s research methodology needs to revised. Unfortunately, we do not know. I also feel that the author is amiss in not referencing prior works that have managed to identify Skype traffic. So here again we do not know whether such identification is enough to block Skype traffic. All in all, this is not an informative analysis.
Posted by aswath at April 20, 2006 04:42 PM
what about a back door embedded within the app itself.. is it feasible ?? :)-
Posted by: /pd at April 21, 2006 04:10 PMSkype embedding detectability? That is funny. :-)
Probably you meant to comment on the previous post on CALEA. LEAs will not accept because the target will discern that intercept is going on. The backdoor can not even be placed in the supernode because the supernode's owner can figure out that some form of intercept going on.
Posted by: Aswath at April 21, 2006 05:29 PM