Studies On Copyright Protection Of MPEG-4 Audio

There will be more and more digital audio applications, with the development of 3G/4G wireless networking. MPEG-4 audio will become the main audio coding standard of these applications, because of its freedom providing for the network end users. Like as other type of multimedia, the copyright protection and multimedia authentication problems must be taken into account during its aimed Internet or mobile applications. This dissertation aims at some key problems, named multi-scheme, multi-watermark issues, of digital watermarking techniques and copyright protection framework for MPEG-4 audio. The works and efforts that the author has done are briefly described as follows.(1) Spread-spectrum audio watermarking in time domain: Modern digital watermarking technology has already experienced the development more then ten years, and spread-spectrum watermarking occupies the significant position during this evolution. Based on the detailed discussion about the algorithms of watermark embedding and detection, asymmetric spread spectrum watermarking and a public detect spread spectrum watermarking schemes, we present a spread spectrum audio watermarking in the time domain, which includes redundancy synchronization to resist some attacks, psychoacoustic model or linear predictive analysis to render the watermark signal imperceptible.(2) Quantization watermarking schemes for MPEG-4 Audio: MPEG-4 Audio integrates many different types of audio coding. As far as we know, there are no current watermarking methods can be fit for all kinds of coding tools, to define the most suitable watermarking scheme for each of them according to its characteristics is the basic assignment research in copyright protection of MPEG-4 Audio. This problem is named multi-scheme coexistence.In this dissertation, the Quantizaion Watermarking (QW) extendedly refers to all the watermarking schemes that achieve watermark embedding during quantization process. Our research indicates that QW can be modeled as digital baseband data communication. The specific designs for MPEG-4 natural sound (music and speech) coding and some theoretic analyses about its rate-distortion-robustness performance demonstrate fundamental principles behind our proposed watermarking technique.(3) Multiple watermarking system: The same MPEG-4 audio coding stream may be divided into more than one audio object, this leads to the multi-ownership identification scenario, since different object could belongs to distinct producers. And at the same time, many actors are involved in the Internet distribution of MPEG-4 audio, such as composer, artist, content provider, and subscriber. In order to protect the rights and interests of all actors...