A Multi-Threshold Based Scheme for Audio Steganography

Audio steganography deals with a method to hide a secret message in an audio file. Also, Audio steganography can be used for secret watermarking or concealing ownership or copyright information in the audio that can be verified later to justify ownership right. A direct Least Significant Bit (LSB) substitution method is one of the most simple and popular techniques used for audio steganography; however, the drawback of this technique is that it can cause hissing sound when the stego audio is played. On the other hand, Human Audio System (HAS) is highly sensitive that can detect or discriminate subtle changes in an audio. In order to improve robustness against hissing noise as well as to avoid detection by a human being, we present a threshold-based scheme for an audio steganography that can selectively embed bits in audio samples based on their magnitudes. Specifically, the scheme can be used to embed zero or many secret bits in an audio sample depending on the relative strength of the audio sample as the sample is measured against a scale of multi-level threshold value. The scheme selects threshold values by analyzing the audio file for a given secret message so that the overall error can be as small as possible while fully utilizing the capacity of the audio file.