A Copyright Authentication Method Balancing Watermark Robustness and Data Distortion

Abstract

Database watermarking plays an irreplaceable role in copyright authentication and data integrity protection, but the robustness of the watermark and the resulting data distortion are a pair of contradictory objects that cannot be ignored. To solve this problem, a reversible database watermarking method, named IGADEW, is proposed to balance the relationship between them. The biggest difference from previous research is that IGADEW synthesizes the optimization objects and obtain various parameters through genetic algorithm (GA). Second, the fitness function considers the weights of robustness and distortion, aiming to find the optimal balance between the two. IGADEW uses the Hash-based Message Authentication Code (HMAC) algorithm to encrypt the experimental parameters and uses the primary key hash algorithm for data grouping, both to ensure robustness. And the data distortion is limited with the help of threshold constraints. Finally, experiments using the UCI dataset demonstrate the effectiveness of IGADEW. Experimental results show that, compared with existing methods, IGADEW is more robust against common attacks, with lower data distortion.