Efficient Anti-Scalable Anti-Collusion Codes

Multimedia content protection technology protects multimedia data against the threats coming from unauthorized users, especially in wired or wireless network environment. The advance of content distribution networks poses a serious threat to content protection against illegal copying and redistribution. This has adverse effects in commercial markets, social people and government security. Generally, protected properties include the confidentiality, integrity, ownership, and so forth. The confidentiality defines that only the authorized users can access the multimedia content, while others cannot know multimedia content. The integrity tells whether media data are modified or not. The ownership shows media data’s owner information that is used to authenticate or trace the distributor.During the past decades, various techniques have been reported for secure multimedia content such as key management, authentication, secure data mining, access control and digital rights management. Although these techniques are able to protect multimedia content’s confidentiality, integrity, ownership; traitor traceability, some of them also suffer from a number of limitations. Indeed, encryption techniques play an important role in content protection, but they are no longer useful after contents are decrypted. To deal with this issue, the idea of fingerprinting and watermarking for traitor tracing has been proposed in recent years.Digital fingerprinting is an emerging technology to protect multimedia content from such illicit redistribution by uniquely marking every copy of the content distributed to each user. Digital watermarking was proposed as a new alternative method to enforce intellectual property rights and protect digital media from piracy. Digital fingerprinting is similar to digital watermarking, except that different information such as the user ID is embedded in each distributed digital contents.However, a general weakness of digital fingerprinting occurs when a coalition of pirates compares their uniquely fingerprinted multimedia to exploit the differences amongst their unique fingerprints in hopes of detecting, removing, or altering the fingerprint so as to evade being traced. This attack is known as collusion. Collusion has been the main research challenge in multimedia security. Indeed, in order to deal with the threat of collusion attacks, Trappe et al. proposed the AND anti-collusion (AND-AC) code against the average attack. Meanwhile, the scheme is vulnerable to linear combination collusion attack (LCCA) and could not support a large number of users. For this last issue, Seol and Kim highlighted SK code by extending the AND-AC code. But SK code is weak against majority attack and linear combination collusion attack. Therefore, the main contribution of this thesis is to design the scalable AC code which is robust to the collusion attack, particularly to the average attack.Indeed, we improve SK scheme by adding to each group a group subcode (GSC). The new code can resist average attack when the inter-group collusion is less likely than intra-group collusion. The collusion attack was modeled as averaging different copies followed by an additive noise. Thus, after determining the guilty groups by using probability method, groups where colluders are coming from, a non-blind detection statistics with the knowledge of the host and soft-threshold detection are used to identify colluders within each guilty group.The proposed model increases the probability of tracing O(nlog-1(N)) colluders within each guilty group, where N is the code length and n the number of users in each group. Experiment results on the real images (lena and cameraman) show that our code is robust to average collusion attack.