An audio self-recovery scheme that is robust to discordant size content replacement attack

Abstract

One of the more severe attacks for self-recovery schemes is the discordant size content replacement attack. This attack substitutes a set of samples from an audio signal with another set of samples of different sizes. In this paper, a watermarking scheme for digital audio self-recovery that is robust to the discordant size content replacement attack is proposed. Our goal is to recover a tampered audio signal after such an attack. In particular, our scheme models the problem using source-channel coding. This method generates a watermarking where a lossy compressor is applied and the source coding output is protected with a channel coder. The coded data is then distributed into the original audio signal. In the recovery stage, a synchronization strategy is used to synchronize the watermark and this makes audio signal self-recovery possible. Our results shows that the propose scheme is robust to the discordant size content replacement attack for attacks of about 20% of the whole audio.