Research On Digital Watermarking Technology For Relational Dat

Relational database technology has an indispensable position in storing,sharing and transacting data.With the rapid development of the Internet,data sharing has become easier and easier.However,the increasing problem of data leakage and tampering in the process of relational data sharing has caused huge losses to enterprises and individuals,thus giving rise to the need for relational database copyright protection.Digital watermarking technology for relational data can help track the source of data,monitor the process of data dissemination,and trace the responsibility in case of data leakage,thus effectively protecting the copyright of data.However,there are still several problems with existing digital watermarking schemes for relational data:(1)Current watermarking schemes have difficulty in simultaneously improving robustness and embedding capacity while maintaining data availability.(2)Digital watermarking schemes that rely on attribute modification inevitably lead to data distortion,which limits the application of these schemes in areas where highly accurate data is required.Although reversible digital watermarking schemes can remove watermark information from data and recover the original data,such data recovery leads to loss of copyright proof for database owners.To address the problem that both watermarking capacity and robustness of relational data watermarking schemes are difficult to improve simultaneously,we proposes a reversible digital watermarking scheme for relational databases based on random embedding.The scheme uses LSB modification method to modify specific bits of Least Significant Bit in an attribute and embeds the specific bits in the LSB bits of the attribute after combining the specific bits with the watermark information using a dissimilarity operation.We have designed watermark embedding algorithm,watermark integrity checking algorithm,watermark detection algorithm and data recovery algorithm.This improves the robustness of watermarking and increases the watermark capacity without increasing the number of modified tuples.During watermark extraction,the attacked bits are discarded to improve the accuracy of watermark detection.To address the distortion problem that watermarking can cause to relational data,we proposes a robust lossless digital watermarking scheme for relational data,and designs watermark encoding and decoding algorithms.The scheme does not make any modification to the original data,thus solving the distortion problem caused by robust digital watermarking.The watermark encoding algorithm combines the bits in the data with the watermark information to generate the auxiliary data.Since there is no need to consider the data distortion problem caused by traditional digital watermarking,the watermarking capacity is significantly increased and the robustness of watermarking is improved with the increase of watermarking capacity.