Database Watermarking Researches Based On Chaos And Hash Function

With the rapid development of computer networks, as well as a wide range ofapplication and development of database management systems and databasetechnology, more and more data are stored and released over the network database.After years of accumulation, these data tend to have some commercial and researchvalue. When these data are treated as product, it will inevitably face some illegalbehavior, such as illegal copying, tampering, and illegal distribution and other issues.If the user obtain the original versions of the database, it is difficult to ensure theownership of the original data set. The use of digital watermark is introduced into thecopyright protection database, because it has the advantages of embedding thecopyright information into a vector product to protect the host product. Databasewatermarking technology can effectively indicate ownership of the database, thusensuring the legitimate interests of the data owner. So database watermarking as aneffective means of copyright protection has become a new application and a new hotspot at current database security.In this paper, it has a series of studies based on digital watermarking technologyin the database. In the First three chapters, it introduces the basic theories relateddigital watermarking and database watermarking, including concepts of digitalwatermarking and watermark database, current research, applications, general systemmodel of database watermarking and some common database watermarkingalgorithm; moreover, chaos theory and hash function are introduced, and they can beused in the database watermarking algorithm as their characteristics and technicalsupport which the database watermarking requires.Based on the theoretical knowledge, the fourth chapter presents a new databasewatermarking algorithm based on chaos and hash function which is against theshortcomings of other algorithms, for example, it cannot be embedded meaningfulwatermark, the number of the vector data of embedding watermark is large, and it hasa greater impact on data. Firstly, it encrypt copyright information using chaotic system, then the resulting binary sequence as to embed watermark sequence toimprove the security of the watermark information. During embedding, the use of twohash function to improve the robustness of the embedded watermark. The first hashvalue is used to filter tuple. The second hash value is used to partition the tuples ofthe candidate tuples, and embedded watermark in every set. In watermark extraction,the watermark error control use the introduction of majority voting method, to ensurethe correct watermark extraction rate, to achieve the blind watermark extraction andblind detection. Simulation results show that the algorithm can not only ensuredatabase availability after watermarking, and when subject to a variety of attack, suchas adding subsets, delete subsets, and modify subsets, it can also ensure the correctwatermark extraction, improved the robustness of the database watermark.