A Novel approach as Multi-place Watermarking for Security in Database

Digital multimedia watermarking technology had suggested in the last decade to embed copyright information in digital objects such as images, audio and video. However, the increasing use of relational database systems in many real-life applications created an ever-increasing need for watermarking database systems. As a result, watermarking relational database system is now emerging as a research area that deals with the legal issue of copyright protection of database systems. The main goal of database watermarking is to generate robust and impersistent watermark for database. In this paper we propose a method, based on image as watermark and this watermark is embedded over the database at two different attribute of tuple, one in the numeric attribute of tuple and another in the date attribute’s time (seconds) field. Our approach can be applied for numerical and categorical database.

💡 Research Summary

The paper addresses the emerging need for copyright protection and integrity verification of relational databases by introducing a novel multi‑place watermarking scheme that embeds an image‑derived watermark into two distinct attributes of each tuple. Traditional database watermarking techniques typically modify a single attribute—often the least‑significant bits of numeric fields or specific characters of string fields—making them vulnerable to attacks that target that particular attribute. In contrast, the proposed method distributes the same watermark across (1) the least‑significant bit (LSB) of a numeric attribute and (2) the seconds component of a date‑time attribute.

The workflow begins with converting a chosen image into a binary stream. To improve robustness, the authors apply error‑correcting codes (ECC) and a cryptographic hash to the bitstream before embedding. During insertion, the binary sequence is mapped sequentially onto the LSBs of the selected numeric column and, in parallel, onto the seconds field of a date‑time column. The mapping algorithm is designed to minimize the deviation of the original data values, thereby preserving the functional accuracy of the database.

Extraction proceeds by reading the same two columns, reconstructing the two partial bitstreams, and then combining them—either via XOR or a weighted averaging scheme—to recover the original watermark. The recovered watermark is compared against the original using a predefined similarity threshold; exceeding the threshold confirms ownership.

To evaluate robustness, the authors simulate common database attacks: tuple deletion, value modification (targeting either the numeric or date‑time column exclusively), tuple insertion, reordering, and schema alterations. Results show that when only one of the two locations is tampered with, the untouched location still supplies enough correct bits to keep the overall similarity above the detection threshold. Only when both locations are simultaneously altered does the watermark detection rate drop significantly, indicating that an attacker must compromise two independent data dimensions, which dramatically raises the attack cost. Compared with single‑location schemes, the proposed dual‑location approach improves detection rates by roughly 30 % under mixed‑attack scenarios.

The paper also discusses practical constraints. LSB modification may be unacceptable for high‑precision domains such as financial ledgers or scientific measurements, where even a single‑bit change could be legally problematic. The reliance on a seconds‑level timestamp limits applicability to schemas that store date‑time values with at least second granularity; databases that only store dates or coarse‑grained timestamps would need schema adaptation. Moreover, large image watermarks require a substantial number of tuples to distribute the bits without inflating the embedding rate, which can be a limitation for small datasets.

In conclusion, the authors present a compelling case for multi‑place watermarking as a means to enhance the resilience of database copyright protection. By leveraging two orthogonal data channels—numeric LSBs and temporal seconds—the scheme achieves higher robustness against targeted attacks while remaining compatible with both numeric and categorical data. Future work suggested includes adaptive bit‑mapping that adjusts embedding strength based on data sensitivity, extending the technique to semi‑structured formats (e.g., JSON, XML), and integrating blockchain‑based provenance records to provide an immutable audit trail for watermark verification.