An Enhanced Hierarchical Traitor Tracing Scheme Based on Clustering Algorithms

Abstract

The easiness of using and manipulating digital media content has a volte face. In fact, although average users can simply be familiar with some manipulations such as a simple duplication, these manipulations can be dangerous with dishonest users whose target is illegal. Manipulating and duplicating digital media content via the Internet and Peer to Peer networks is available even to average users but can be used to unauthorized purposes with dishonest customers. Henceforth, facing the loss caused by unauthorized treatments and protecting the digital content become challenging to the media industry and research has led to different mechanisms of digital content protection. The aim of the multimedia distribution platforms, even Video on demand platforms, is to propose a suitable structure to the embedded fingerprints to ensure an efficient and fast tracing process in multimedia distribution platforms involving great number of users. The Tardos code has been the most popular tracing code due to its efficient tracing detection performance. One main challenge of the existing Tardos-based tracing approaches was to face the decoding complexity and the computational costs of the tracing process.

Hence, the tracing scheme we propose to improve in this paper was proposed previously as a group-based scheme which enables to construct groups of users according to a multi-level hierarchy. Based on clustering algorithm, we propose to construct groups of users’ fingerprints, and then to apply the tracing process. The main target is to show how deep is the impact of using a clustering algorithms in the hierarchical tracing scheme.