An inaudible pseudonoise sequence generation for digital audio watermarking

Digital watermarking is a technique that embeds imperceptible and statistically undetectable information into digital data. Most digital audio watermarking schemes have used pseudonoise (PN) sequence as a secret key. PN sequence embedded in an audio signal for watermarking reduces audio quality while it plays a vital role in security. In this paper, a PN sequence generation to improve imperceptibility in perceptual region of the human auditory system for audio watermarking purpose. The basic idea of this work was a modification of a PN sequence according to the frequency analysis of adjacent binary digits that have different signs in a PN sequence. The canonical signed digit code (CSDC) was also used to modify a general PN sequence and the pair-matching function between general and its modified PN sequence is proposed for PN sequence modification. Additionally, statistical properties of new PN sequence was evaluated and discussed by an auto-correlation and a cross-correlation analysis.