On the effects of pirate evolution on the design of digital content distribution systems

  • Authors:
  • Aggelos Kiayias

  • Affiliations:
  • Computer Science and Engineering, University of Connecticut, Storrs, CT

  • Venue:
  • IWCC'11 Proceedings of the Third international conference on Coding and cryptology
  • Year:
  • 2011

Quantified Score

Hi-index 0.00

Visualization

Abstract

A cryptographic primitive that is widely deployed commercially for digital content distribution is the subset-difference (SD) method of Naor, Naor and Lotspiech that was introduced in Crypto 2001. This encryption mechanism, called a trace and revoke scheme, is part of the Advanced Access Content System (AACS), and is used for encrypting Blu-Ray movie disks and is based on an explicit combinatorial construction of an exclusive set system. At the time of its introduction the only attacks cryptographers considered against such schemes were against the revocation and tracing algorithms. The SD method defended against them successfully and provided a superior ciphertext length compared to other known techniques : the length of the ciphertext grew only linearly with the number of revocations r; in contrast, e.g., the simpler complete subtree (CS) method requires ciphertexts of length O(r ċ log N/r) where N is the total number of users. In Crypto 2007 a new class of attacks was discovered against trace and revoke schemes called "pirate evolution." Pirate evolution refers to the ability of the adversary to schedule the key material it possesses in such a way so that it can withstand a great number of rounds of tracing and revocation. With the introduction of pirate evolution, the reduction of the number of rounds of pirate evolution became a design consideration for trace and revoke schemes. In 2009, Jin and Lotspiech proposed a mechanism for defending against pirate evolution in the SD method that is a tradeoff between ciphertext size and the pirate evolution bound. In this article we provide a review of all the above results. Moreover, we compare the modified SD scheme to the CS method (similarly modified to address pirate evolution) and find that for many choices of the parameters that are relevant to practice SD can be a less preferable choice. This fact highlights the importance of considering all relevant attack scenarios when applying a specific cryptographic primitive to a certain application domain.