Research On Robustness Digital Audio Watermarking Algorithm For Resistance To Synchronization Attack

Digital audio watermarking is a new technology for copyright protection,authentication and the integrity of the source under an open network environment. Robust digital audio watermarking is mainly used to logo copyright information in the digital works.An algorithm is implemented in this paper. The algorithm researches the robustness of digital audio watermarking with anti-synchronized attacks in transform domain. Its main purpose is to resist various synchronize attacks with high-intensity. Synchronization attacks have more resistance than simple ones, because synchronization attacks will destroy the information-bearing data and the watermark synchronization. Watermarking still exists when digital works have been attacked. There is no change in magnitude, but the watermark signal has been misplaced and can not maintain normal synchronization which is required in the process of watermark extraction. Two algorithms are researched. The first one is to study robust digital audio watermarking algorithms with anti-synchronized attacks by using DWT. The embedding strength of watermark is adjusted adaptively according to SNR. Combining with the auditory characteristics of human, the coefficients from the watermark signal and the original audio signal, which are implemented separately by DWT, can be better integration. It takes three kinds of different audio signals as carrier. The common attack and the synchronized attack are made separately on to the watermark. Through many simulation experiments, it shows that the algorithm can enhance the performance of unperceived in watermark. The second one is based on the first algorithm. The multi-resolution of DWT is combined by the de-correlation capability and energy aggregation of DCT. At the same time, the synchronization technology is introduced in the audio section. The blind detection of watermark can be achieved under the guarantee at a good balance of unperceived and robustness. The technique for correct error coding is used to embed watermark in order to reduce detection error rate of watermark. Blind extraction is implemented at the receiver. Experimental results show that this algorithm has more perfect performance of robustness and unperceived than anti-synchronized attacks used for DWT. The algorithm performance is improved.