A Privacy Protection-Capable Digital Fingerprinting Scheme with Semi-Blind Detection for Aggregated Data

Abstract

Digital fingerprinting tracks traitors who illegally distribute digital products by creating a unique code for each user. Existing digital fingerprinting schemes are more concerned with designing more powerful and efficient fingerprint encoding and detection algorithms, yet few of them pay attention to the privacy protection of the digital carrier itself. This paper presents a digital fingerprinting scheme with privacy protection capability, which can simultaneously protect the privacy of data and track traitors by fusing the differential privacy and digital fingerprinting technology. And the proposed Privacy Protection-capable FingerPrinting scheme (PPFP) embeds an elaborately designed noise set at specific positions in a large data set, realizing fingerprint embedding and noise interference in one step. In our PPFP scheme, since the identifying fingerprint information is encoded as the coordinate point with embedded noise in the data set, the semi-blind detection can be achieved by comparing the hash values of the original data set and the illegal data set to obtain the coordinates with inconsistent content. The PPFP scheme can not only guarantee the privacy of the Traceable Noised Data set (TND) generated by it, but also have the basic characteristics of digital fingerprinting, such as imperceptibility, robustness, credibility and feasibility. The analyses are achieved by formula derivation and formal language verification, while simulation and the security attack experiments are performed using real data. The analysis and simulation results show that the PPFP scheme is sufficient to meet the requirements of privacy-enhanced traceability.

Keywords: Digital fingerprinting, Differential Privacy, Privacy protection, traitor tracing, semi-blind detection