基于随机性统计特征的隐密分析研究

Study of Randomness-based Statistical Steganalysis

隐密分析技术作为检验隐密术性能和挫败非法信息传播的重要手段,其研究日益受到人们的关注。通过对主要隐密分析方法的研究与比较,指出以随机性统计特征为基础的统计攻击是目前的主流,称为RSS分析。描述了RSS分析的基本假设,即加密压缩后的信息具有比载体隐密区域更强的随机性。从基本条件、统计特征、判决规则、隐密信息长度估计等方面分析了一些具有代表性的RSS分析的原理及其各自的优缺点。在此基础上讨论了RSS分析的基本假设普遍存在的问题,并指出应从理论研究和方法创新等方面进行进一步的研究。

As the important test means of the performance of steganography,steganalysis can also be used to frustrate the illegal information distribution.By the analysis of the prevalent methods of steganalysis,we find out that they are almost all based on some random statistical features.We call them as the randomness-based statistical steganalysis,namely the RSS.The basic hypothesis of RSS methods is presented at first that the encrypted or compressed data is more random than the data of covers.Then the principles and shortcomings of some typical RSS methods are proposed in aspects of the basic condition,the statistical feature,the judgment rule and the hided data length estimation.Based on the above analysis,the disadvantages of the basic hypothesis of RSS are discussed.At last,according to the RSS,the theory research and the innovations are also pointed out.