Research on digital audio watermark technology for broadcast information security

This paper introduces the technology of spread-spectrum watermarking of audio signals. We adopt the direct-sequence spread-spectrum watermark, which is based on rate concept, making use of chess watermark to shape watermark signals. After making audio signals MCLT transform, according to the Psycho-Acoustic Frequency Masking (PAFM), we use multiplication to add the watermark to the MCLT transform domain of audio signals. The added location is determined by pre-detection technology. In the watermark detection side, we use the cepstrum filter to reduce the variance of the correlation test to improve the watermark detection reliability. This paper proposes to change both the spread factor and the repeated factor to reduce window size and enhance watermark capacity considering the robustness, imperceptibility, and watermark capacity. The system has certain practical and referenced value, and is a useful exploration for the utility of broadcast monitoring technology in the future, which is with inspiration and reference.