High Capacity and Secure Temporal Audio Steganography Model with Robustness Against Amplification

Temporal domain audio steganography methods have clear benefits of being simple and computationally quick. However, the existing temporal domain methods that have high capacity, like the least significant bit (LSB) method, are highly sensitive to amplification attacks, which are very easy to implement. Amplification can completely destroy the hidden message even at undetectable levels such that discerning an attack from unintentional damage is hard. On the other hand the temporal domain algorithms that are actually robust against amplification have very low hiding capacity. A new method of temporal audio steganography is proposed here which is robust against amplification attacks while having high capacity, security and perceptual transparency. It builds on the basic difference expansion paradigm and applies cryptographic principles for enhanced security. The algorithm takes a stereo WAV cover file and secret message file, compresses and encrypts the secret file, and employs a transform domain algorithm to embed the resulting data into the cover WAV file. By exploiting properties of stereo WAV files, the algorithm succeeds in being more robust toward attacks of amplification. Performance analysis is done using peak signal to noise ratio (PSNR), signal to noise ratio (SNR), mean square error (MSE) and Hamming distance.