DPC

DPC: Availability / content protection

Person in charge: Caroline Fontaine (mail)

Pedagogical team: Gwenaël Doërr, Caroline Fontaine, Mohamed Karroumi (content protection), Eric Totel, Frederic Tronel (availability) (global mail)

Description

This course is composed of two distinct parts, each representing 10h teaching: the first one deals with systems and applications availability of systems, while the second one is about content protection (access control, digital rights management, traitor tracing).

Availability of systems

This lecture addresses the problem of the availability, one of the major attributes of dependability. The stress is laid on the design and the use of redundancy mechanisms in the presence of transient faults. These kind of faults can be either accidental or intentional.

The main part of the course is devoted to the study of software solutions that ensure the availability of a critical service accessed by external entities. Fault tolerant mechanisms are analyzed from the points of view of their architectures as well as their fundamental algorithmic solutions.

Content protection

This part presents tools that enable to manage (multimedia) content diffusion, as we want to control their access, make easier and more powerful the copyright protection and the management of digital rights, or be able to face piracy and trace traitors.

Keywords: access control, digital rights management, trator tracing, piracy (content protection part)

Prerequisites: some basic knowledge about algorithmic of the reliable distributed systems (availability), basics of cryptography (encryption, signature, protocols) (content protection part)

Learning outcomes

• Availability, replication, diversification distributed algorithms
• Design of dependable systems
• Content protection of media, DRM, digital watermarking, traitor tracing
• Design and analysis of content protection systems

Contents

• Availability of systems (10h)
  • Definitions and Architectures of the Solutions (4h) : After a recall of the concepts related to reliability and availability, we consider first accidental faults (crashes and errors of design) and then intentional faults (arbitrary behaviors). We present some mechanisms that ensure fault tolerance (detection, covering, maskings) and then mechanisms that allow to detect (functional diversification) and to tolerate intrusions (fragmentation-redundancy-dissemination).
  • Calculability in Unreliable Asynchronous Environments (6h) : All the solutions described previously have to maintain consistency among a set of distributed replicas. This problematic has been studied for many decades. To solve these problems, it is necessary to provide solutions to various agreement problems. The literature abounds with theoritical results about the solvability/unsolvability of these problems with respect to the hypothesis that are made about the considered system. When one restricts to asynchronous environments, most of these results are about impossibility to solve these agreement problems. We will clarify these results and we will illustrate the strategies that have been adopted to circumvent them (by increasing the assumptions about the model or by weakening the specification of the problem). A study of the transitory arbitrary faults will offer the opportunity to identify the relations that can exist between mechanisms ensuring the availability and recovery mechanisms.
• Content protection (10h)
  • Content protection (2h) : this first part will introduce the concepts of copy protection and some basic techniques for content protection, taking into account the more recent solutions used in the area of digital cinema or superdistribution; some nowadays standards will also be introduced.
  • Digital Rights Management (3h) : this part will go deeper in the three theoretical and practical levels of any modern DRM system: how to express the rights (XrML, ODRL), usage rights application, and content protection. The OMA DRM and Marlin systems will be more precisely presented. At last, the inter-operability property for such systems (Coral, DVB, DMP) will be discussed.
  • Digital watermarking (2h) : the scope of this small of the course is to briefly present the context of digital watermarking, as a complement of the more classical DRM techniques, embedding in the medium an imperceptible signal that will help attest the identity of the document owner or user, or simply transmit some metadata to increase its potential. Student will not need any sgnal processing knowledge to attend this introduction.
  • Traitor tracing (3h) : this last part will beal with the ability to trace piracy when a medium is delivered to several people. It takes place in the context of the “broadcast encryption”, or superdistribution. The idea is to design the scheme in a way such that if several user collide to produce a fake decoder (or a fake copy of the medium), we will be able to identify at least one of them while accessing this fake decoder/copy. The practical example of AACS (designed for HD DVD) will be presented.

Bibliography

• Arlat J., Blanquart J.-P., Costes A., Crouzet Y., Deswarte Y., Fabre J.-C ., Guillermain H., Kaâniche M., Kanoun K., Laprie J.-C. J.-C, Mazet J.-C.,
• Powell D., Rabéjac C., Thévenod P. Guide de la sûreté de fonctionnement. Laboratoire d'ingénierie de la sûreté de fonctionnement (LIS), Cépaduès - Éditions, 1995
• Davoine F., Pateux S. Tatouage de documents audiovisuels numériques. Hermès-Lavoisier, 2004
• Furht B., Kirowski D. Multimedia security handbook. CRC Press, 2004

Evaluation mode

Evaluation is based on quizzes, reviews, and/or other academic exercises mostly outside scheduled class times.
The final grade is determined by performance on:

• quizz (1/3), oral presentation of a research paper (2/3)