Spread-spectrum steganalysis and PN sequence estimation

A novel audio steganalysis approach for Direct-Sequence Spread-Spectrum (DSSS) hiding method is presented in this paper. The proposed scheme can potently detect DSSS hiding message without any priori knowledge. Meanwhile, it can give an optimum estimation of the Spread Spectrum (SS) sequence adopted by transmitter. With a given length of PseudoNoise (PN) sequence, the maximal cumulative cross correlation between PN sequence and test audio can be obtained employing Genetic Algorithm (GA), and then the kurtosis of cross correlation curve via PN sequence length is adopted as the sensitive feature to design the detector. The optimum estimation of SS sequence is considered as the PN sequence which makes the curve achieve its maximal value. Experiments have been carried out under various embedding rates, and the results demonstrate the effectiveness of the proposed scheme.