Open Access

ARTICLE

A Novel Database Watermarking Technique Using Blockchain as Trusted Third Party

Ahmed S. Alghamdi¹, Surayya Naz², Ammar Saeed³, Eesa Al Solami¹, Muhammad Kamran^1,*, Mohammed Saeed Alkatheiri¹

1 Department of Cybersecurity, College of Computer Science and Engineering, University of Jeddah, KSA
2 Abdul Wali Khan University, Mardan, Pakistan
3 Department of Computer Science, COMSATS University Islamabad, Wah Campus, Wah Cantt, Pakistan

* Corresponding Author: Muhammad Kamran. Email:

Computers, Materials & Continua 2022, 70(1), 1585-1601. https://doi.org/10.32604/cmc.2022.019936

Received 02 May 2021; Accepted 03 June 2021; Issue published 07 September 2021

Abstract

With widespread use of relational database in various real-life applications, maintaining integrity and providing copyright protection is gaining keen interest of the researchers. For this purpose, watermarking has been used for quite a long time. Watermarking requires the role of trusted third party and a mechanism to extract digital signatures (watermark) to prove the ownership of the data under dispute. This is often inefficient as lots of processing is required. Moreover, certain malicious attacks, like additive attacks, can give rise to a situation when more than one parties can claim the ownership of the same data by inserting and detecting their own set of watermarks from the same data. To solve this problem, we propose to use blockchain technology—as trusted third party—along with watermarking for providing a means of rights protection of relational databases. Using blockchain for writing the copyright information alongside watermarking helps to secure the watermark as changing the blockchain is very difficult. This way, we combined the resilience of our watermarking scheme and the strength of blockchain technology—for protecting the digital rights information from alteration—to design and implement a robust scheme for digital right protection of relational databases. Moreover, we also discuss how the proposed scheme can also be used for version control. The proposed technique works with nonnumeric features of relational database and does not target only selected tuple or portion (subset) from the database for watermark embedding unlike most of the existing techniques; as a result, the chances of subset selection containing no watermark decrease automatically. The proposed technique employs zero-watermarking approach and hence no intentional error (watermark) is added to the original dataset. The results of the experiments proved the effectiveness of the proposed scheme.

---

Keywords

---