Digital forensics in VoIP networks

With VoIP being deployed on large scale, forensic analysis of captured VoIP traffic is of major practical interest. In this paper, we present a new fingerprinting approach that identifies the types of devices (name, version, brand, series) in captured VoIP traffic. We focus only on the signaling plane and discard voice related data. Although we consider only one signaling protocol for the illustration, our tool relies on structural information trees and can easily be adapted to any protocol of that has a known syntax. We have integrated our tool within the well known tshark application in order to provide an easy to use support for forensic analysts.