论文部分内容阅读
S盒是对称密码算法中的重要组成部分,作为主要的非线性部件,其密码性质的好坏直接影响到整体算法的安全性.差分均匀度和线性度(非线性度)是衡量S盒密码性质的两个基本指标,它们分别刻画了S盒抵抗差分密码分析和线性密码分析的能力,并且在仿射变换下保持不变.由于硬件成本限制,轻量密码算法通常采用4比特S盒,其差分均匀度和线性度的下界为4,达到下界的S盒称为最优S盒,Leander等将它们分成了16个仿射等价类.在此基础上,我们对现有典型轻量算法中的S盒按仿射等价关系进行了分类.为了对抗多差分分析、多线性分析及各种变形攻击方法的威胁,还希望S盒具有最大差分概率的差分对个数、具有最优线性逼近关系的掩码个数越少越好,有时甚至需要对单比特输入输出的差分特征和线性特征做更细致的分析,因此我们进一步对上述各轻量S盒达最大差分概率的差分对个数、具有最优线性逼近关系的掩码个数、单比特输入输出差分特征和单比特线性逼近关系的个数,以及单比特情况下的差分均匀度和线性度进行了详细的分析和统计,上述结论可为相关轻量密码算法的分析提供重要的理论依据.
S-box is an important part of the symmetric cryptography algorithm, as the main non-linear component, its cryptographic properties directly affect the security of the overall algorithm.Difference uniformity and linearity (non-linearity) is a measure of S-box password Nature of the two basic indicators, which respectively characterize the S-box resistance to differential cryptanalysis and linear cryptanalysis capabilities, and affine transformation remains unchanged due to the hardware cost constraints, lightweight cryptographic algorithms usually use 4-bit S-box, The lower bound of differential uniformity and linearity is 4, the S box reaching the lower bound is called the optimal S box, and Leander et al divide them into 16 affine equivalence classes.On this basis, we analyze the existing typical lightweight The S boxes in the algorithm are classified according to their affine equivalence relations.In order to counter the threats of multi-difference analysis, multi-linear analysis and various deformation attack methods, it is also hoped that the number of differential pairs with S-boxes with the largest difference probability is optimal As the number of masks in the linear approximation relationship is as small as possible, and sometimes even a more detailed analysis of the differential characteristics and the linear characteristics of a single bit input / output is required, we further extend the maximum difference probability The number of differential pairs, the number of masks with the optimal linear approximation, the number of single-bit input-output differential characteristics and the number of single-bit linear approximations, and the evenness and linearity under single-bit conditions are analyzed in detail And statistics, the above conclusions can provide important theoretical basis for the analysis of the relevant lightweight cryptographic algorithms.