Progressive audio scrambling via wavelet transform

Audio scrambling methods are widely used for providing confidentiality in audio distribution. In this paper, an effective audio scrambling method based on a pseudorandom number generation algorithm, namely Mersenne Twister, is firstly proposed. Then, with the aim of strengthening anti-decryption capability, that method is applied in wavelet domain. According to experimental results, the proposed method is fast, simple and has strong anti-decryption/anti-attack capabilities. Moreover, it achieves progressive scrambling in which quality of the scrambled audio files can be controlled based on the system requirement: severely distorted ones for high-level security and slightly distorted ones for preview models of DRM systems.