Statistical Pattern Recognition Based Content Analysis on Encrypted Network: Traffic for the TeamViewer Application

Authors:
  • RARobert Altschaffel
    Robert Altschaffel
  • RCRobert Clausing
    Robert Clausing
  • CKChristian Kraetzer
    Christian Kraetzer
  • THTobias Hoppe
    Tobias Hoppe
  • SKStefan Kiltz
    Stefan Kiltz
  • JDJana Dittmann
    Jana Dittmann

Abstract
In the course of a forensic investigation it might be required to distinguish between different network activities. While various means to analyse network traffic exist, encrypted traffic often makes such an analysis problematic. The focus of this paper is to introduce a method based on statistical pattern recognition on network recordings of encrypted sessions to distinguish between different actions within these sessions. For demonstration purposes the popular remote support and online-meeting application TeamViewer is selected to introduce and discuss an approach to distinguish between file transfers, voice conferences, video conferences, text chat and normal remote sessions within TeamViewer sessions.