Research On Multicast Source Authentication Based On Digital Waremarking Technology

With the rapid development of communication technology and the Internet, multicast technology is well developed, but the resulting that multicast security issues have become increasingly prominent, one of the most problems to be solved is the multicast source authentication problem. The development of digital watermarking technology provides the solution for Multicast Source Authentication. We consider the multicast network features and digital watermarking technology to solve the problem of multicast source authentication. Also we improve the security and recoverability of the digital watermark technology. The main work of this paper is as follows:(1) A multicast source authentication method based on digital watermark technology has been proposed. This method solved the multicast source authentication by improving the current digital watermark method. This method introduces the public key cryptography to solve the problem of authentication. By modifying the feature extraction method, this method completed the blind detection, increase the ability of safety. The simulation experiments tell us that:the multicast source authentication method based on digital watermark technology is active.(2) A digital watermark method based on multi-feature extraction has been proposed. With the watermark generated by multi-feature extraction combines the edge features and DCT coefficients of the characteristics of the relationship as the watermark information. The watermark information can be more fully represent the image content information and overcome authentication failure because of the single feature attacks. The simulation experiments tell us that:under the premise of ensuring the invisibility, this method has a good robustness when it responses to the normal image processing, and it has a good recognition positioning when it responses to the malicious tampering attacks.(3) A digital watermark method based on the relationship coefficient between the block has been proposed. This method introduces the embedding strength which control the embedding strength of each image block to ensure the embedded watermark image is not visible. In addition, we propose a self-embedding watermarking method which uses the coefficient of relationship between the images. From the image block quantization coefficient we extract the contents of the vector. We use the relationship between the two contact image content vector as the watermark information and as the basis for self-recovery of tampered image. This self-embedding watermarking method reduces the watermark embedding capacity. The simulation experiments tell us that:under the premise of ensuring the invisibility, this method improves the quality of the tampered image recovery.