SEMI-FRAGILE WATERMARKING SCHEME FOR RELATIONAL DATABASE TAMPER DETECTION

Most of data over the Internet today is hosted on outsourced third-party servers which are not trusted. Sometimes to be distributed other organizations or individuals for pre-agreed use. In both these scenarios susceptible malicious tampering so there a need some mechanism verify database integrity. Moreover, authentication process should able differentiate between valid updates and modifications. this paper, we present novel semi-fragile watermarking scheme relational integrity verification. Besides detection localization tampering, proposed allows modifications that periodic updates, without requiring re-watermarking. Watermark embedding distortion free, as it done by adjusting text case selected values resulting in retention semantic meaning data. Additionally, group-based ensures up group level. We implemented proof concept application our technique. Theoretical analysis experiments show even single value modification can detected with very high probability besides attacks like tuple insertion deletion.