扩频卷积指纹及其安全性分析  

Spread Spectrum Convolutional Fingerprinting and Its Security Analysis

作　　者：朱岩[1] 韩心慧[1] 叶志远[1] 杨永田[2] 

机构地区：[1]北京大学计算机科学技术研究所,北京100871 [2]哈尔滨工程大学计算机科学与技术学院,哈尔滨150001

出　　处：《计算机学报》2006年第6期960-968,共9页Chinese Journal of Computers

摘　　要：数字指纹是一种在每个数字作品拷贝中添加唯一性信息的版权保护技术.文中证明了扩频序列对组合、平均及噪声攻击具有有效的合谋容忍性质,并提出一种卷积码与扩频水印码相链接的安全指纹方法.译码方法也通过将备选子码集应用于Viterbi译码算法给予改进,进而合谋安全性和性能被分析与证明.结果表明该方法具有更短的码长和更快的叛逆者搜索性能.Digital fingerprinting schemes is a technique applied to protect the copyright by embedding the unique information in each distributed digital contents. This paper proves that spread-spectrum sequence has the effective collusion-resistant properties against malicious combination, average and additive noise attacks. A secure fingerprinting is presented by concatenating the spread-spectrum code with a convolutional code. The decoding method is also improved by using Optional-Code Set to Viterbi algorithm. Moreover, the collusion security and performance are analyzed and proved. The results show that the proposed fingerprinting has both shorter code-length and faster traitor searching.

关 键 词：数字指纹 合谋安全 扩频 卷积编码 

分 类 号：TP309[自动化与计算机技术—计算机系统结构]