Construction and Performance Analysis of Digital Fingerprinting Code with Identifiability of All Malicious Users

数字指纹代码嵌入在已许可的数字内容中，以防止恶意用户非法分发。数字指纹码通常被要求具有识别部分（或全部）恶意用户合谋的能力，本文考虑了数字指纹码的信息论模型，并刻画了用户的最大速率（数字指纹编码的容量）使得在特定假设下，可以以接近1的概率指定共谋中的所有恶意用户。我们还讨论了零错误能力的数字指纹代码的情况下，所有的恶意用户在共谋指定没有错误在某些攻击模型。

Digital fingerprinting codes are embedded in licensed digital contents for preventing illegal distribution by malicious users. Digital fingerprinting codes are usually required to have the ability to specify a part (or all) of a collusion of malicious users who generate the pirated copy.In this study we consider an information-theoretic model of the digital fingerprinting code and characterize the maximum rate of the users (the capacity of the digital fingerprinting code) such that all the malicious users in a collusion can be specified with probality close to one under a certain assumption. We also discuss the zero-error capacity of the digital fingerprinting code for the case where all the malicious users in a collusion are specified without error under certain attack models.