Distortion-free fragile watermark for relational databases. (2016)

• Record Type: Journal Article
• Authors: Kamel, Ibrahim; AlaaEddin, Maha; Yaqub, Waheeb; Kamel, Kareem
• Is Part Of: International journal of big data intelligence. Volume 3:Number 3(2016)
• Page Start: 190
• Page End: 201
• Publication Date: 2016
• Subjects: data integrity -- digital watermarking -- relational databases -- information security -- distortion -- fragile watermarks -- tuples -- simulation -- attack detection; Big data -- Periodicals; 005.705
• Languages: English
• ISSNs: 2053-1389

Abstract

Unauthorised changes to databases may result in significant losses for organisations. Digital watermarking can be used for protecting the integrity of relational databases. Prior techniques introduce distortions to the watermarked attributes and thus limit their application. This paper, on the other hand, proposes a distortion-free watermarking mechanism for protecting the integrity of relational databases. The proposed technique establishes a one-to-one relationship between the secret watermark and the relative order of tuples in a group. Alterations by attackers, who do not know the secret watermark, will disturb the secret order and thus destroy the existing fragile watermark. We carried out extensive simulation experiments to evaluate the effectiveness of the proposed technique. The experiments showed that 100% detection rate can be achieved by increasing the group size. Moreover, if the number of attacks is limited to one, the proposed technique can identify, with high probability, the attacked attribute and the type of attack.