Tardos fingerprinting is better than we thought
Source: Cornell University
Tardos has proposed a randomized fingerprinting code that is provably secure against collusion attacks. The authors revisit his scheme and show that it has significantly better performance than suggested in the original paper. First, they introduce variables in place of Tardos' hard-coded constants and they allow for an independent choice of the desired false positive and false negative error rates. Following through Tardos' proofs with these modifications, they show that the code length can be reduced by more than a factor of two in typical content distribution applications where high false negative rates can be tolerated. Second, they study the statistical properties of the code. Under some reasonable assumptions, the accusation sums can be regarded as Gaussian-distributed stochastic variables.