Research On Audio Data Hiding Using Psychoacoustic Model

Information Hiding is a whole new technology for information security protection. By its means, secret information is hidden into host signal, which can be transmitted and shared in public, with secret information not recognized by the public. This technology is with high academic value and economic value, since it covers the research areas of communications and information theory, cryptography, digital image processing, network security, etc, and have a great many applications in hiding data, copyright protection, secret communications, and so on. One important branch of information hiding technology, audio data hiding, with the feature of using audio as host signal, has become a research focus. In this paper, firstly we give a brief introduction about the background, basic terminology, general model and applications of information hiding technology. Then the principles of the digital audio data hiding are discussed in detail. Furthermore, typical audio data hiding schemes are expatiated, and their advantages and disadvantages are analyzed. To achieve fine imperceptible characteristic, the audio model in widely used audio coding theory is introduced into the audio data hiding research. By analyzing human physiological audio system and its masking effect, this paper illustrates the psychoacoustic model in audio part of the ISO video coding standard MPEG1. A novel audio data hiding algorithm, which utilizes this psychoacoustic model of the host audio signal to regulate secret information, is proposed. In the extraction process, the hidden data can be correctly detected without original host audio signal, which is called blind extraction, because of the advantage of the fine correlation characteristic of m-sequence. Experiments show that the imperceptible characteristic of the embedded audio signal is good and the bit error is low no matter any attack is existed or not. Finally, it's the overall summarization of the paper and research expectations of the digital audio data hiding techniques.