Network based intrusion detection to detect steganographic communication channels: on the example of audio data

Due to the nature of multimedia data exchange over the Internet between authorized users, one of the most difficult tasks is to control unauthorized information flow. With respect to confidentiality requirements and breaches in security, steganography offers an instrument for attackers to hide a communication by embedding the secret message into an innocuous communication channel, the so-called cover data. Due to the improvements of audio steganography, disclosure of information becomes possible for a lot of applications. In our paper we investigate voice-over-IP applications and intrusion detection systems (IDS) to detect hidden communication channels. Besides an general application scenario and performance of audio capturing, we introduce two steganalysis classifier for speech: one for MP3Stego and one for Steghide. Our first tests are based on an offline analysis of captured speech data.