轻量级分组密码算法MIBS不可能差分分析

轻量级分组密码算法适用于RFID等计算资源严格受限的环境。指出之前Asli Bay等人不可能差分分析结果的错误,并提出新的不可能差分分析。提出对约减至12轮的MIBS算法的攻击,此攻击需259选择明文和263次加密运算。     

一类广义Feistel密码的安全性能分析

为评估一类广义Feistel密码的安全性能,通过列举的方法,对该分组密码抵抗差分密码分析和线性密码分析的能力进行了深入研究.在轮函数是双射的假设条件下,证明了4,8,12,16轮广义Feistel密码分别至少有2,5,8,10个轮函数的输入差分非零;证明了4r(r≥2)轮广义Feistel密码至少有2r 1个轮函数的输入差分非零.从而若设轮函数的最大差分和线性特征的概率分别为p和q,则4r(r≥2)轮广义Feistel密码的差分特征和线性特征的概率分别以p2r 1和q2r 1为其上界.     

对MIBS算法的中间相遇攻击

MIBS是一种Feistel结构的轻量级分组密码算法,它广泛适用于RFID等计算资源严格受限的环境.评估了简化轮数的MIBS算法对中间相遇攻击的抵抗能力,研究了MIBS算法的一个等价结构,构造出它的一个6轮区分器.基于6轮区分器,利用中间相遇攻击的方法,分别对8/9/10轮的MIBS-64/80和11轮的MIBS-80进行了攻击,并利用密钥生成算法中轮密钥之间的关系降低了数据复杂度和时间复杂度.所有的攻击实例都给出了复杂度分析.攻击结果表明11轮的MIBS-80对中间相遇攻击是不免疫的.     

改进的44轮SHACAL-2的相关密钥攻击

采用相关密钥的三明治矩形攻击,改进了44轮SHACAL-2的相关密钥攻击,利用模减差分和异或差分的混合表示方式以及采用差分集合代替单个差分提高差分路线的概率,构造的35轮相关密钥三明治矩形区分器的概率为2-430。利用该区分器给出了相关密钥情况下44轮SHACAL-2的密钥恢复攻击,复杂度为2217个选择明文,2476.92次44轮SHACAL-2加密,2222字节存储。     

利用扩散层固定点对ARIA密码攻击的改进

ARIA密码算法是韩国学者提出的韩国分组密码标准,该文对ARIA算法扩散层固定点进行了研究,结合固定点和S盒的差分性质,构造了达到概率上界2-144的6轮差分传递链.此外,利用特殊固定点构造了新的形式为4→4的截断差分路径,实现了7轮截断差分攻击.新的攻击数据和时间复杂度约为2106,存储复杂度约为256.     

Piccolo结构抵抗差分和线性密码分析能力评估

Piccolo结构是从Piccolo算法中归结出来的一种分组密码结构,该结构的特色在于轮函数和移位变换的设计。为评估Piccolo结构的密码性能,对该结构抵抗差分密码分析和线性密码分析的能力进行了研究。给出了任意轮差分特征中活动轮函数和活动S盒个数的一个下界,并通过研究Piccolo结构的差分线性对偶性,进而给出了任意轮线性逼近中活动轮函数和活动S盒个数的一个下界。     

热传导方程的一个高精度8点隐式差分格式

主要考虑一维标准的热传导方程初边值问题的有限差分解法。利用Taylor展开与待定系数的方法,构造出一个2层8点隐式差分格式,并得到该差分格式的局部截断误差为Q(τ~3+h~6)和稳定条件为0r≤0.126或0.203≤r≤1.352。最后,给出了2个数值算例以验证所得差分格式的计算效果。     

Piccolo结构抵抗差分和线性密码分析能力的进一步评估

为评估Piccolo结构的密码性能, 对该结构抵抗差分密码分析和线性密码分析的能力进行研究。给出任意轮差分特征中活动轮函数和活动S盒个数的一个新的下界, 并利用Piccolo结构的差分线性对偶性, 给出任意轮线性逼近中活动轮函数和活动S盒个数的一个新的下界。同时, 证明这些下界是不可改进的。     

基于汉明重的MIBS密码代数旁路攻击

对MIBS密码代数旁路攻击能力进行了评估:首先给出代数旁路攻击模型,构建MIBS密码等价代数方程组;采集微控制器上MIBS密码实现加密过程中功耗泄露,并选取功耗特征明显的部分泄漏点,基于模板分析进行加密中间状态汉明重推断;利用SAT(可满足性)、PBOPT(伪布尔随机优化)及LP(线性编程)分别对MIBS代数方程组和汉明重泄露进行表示;最后采用SAT问题解析器CryptoMinisat和混合整数编程问题(SCIP)解析器进行密钥恢复,并在不同场景下进行大量攻击实验.结果表明:MIBS密码易遭代数旁路攻击;汉明重推断正确条件下,利用1条功耗轨迹中4轮汉明重信息泄露可成功恢复完整密钥;汉明重推断部分正确条件下,基于SAT,PBOPT和LP这3种代数方程组求解方法,可分别在汉明重推断错误率不超50％,65％和60％的情况下成功恢复MIBS完整密钥.     

色散方程八点显格式的稳定性

文中对色散方程z_i=z_给出了一个带参数的八点显式差分格式,还过对参数的全面讨论,得到了稳定的最佳范囹r≤3.154,它比以往的结果(r≤2.39)好了许多。     

Improved differential attack on 30-round SIMON64

In this paper we present an attack on 30-round SIMON64, which improves the best results on SIMON64 by 1 round. We use a 23-round differential characteristic which was proposed by Itai et al in 2015 to construct a 30-round extended differential characteristized by adding 4 rounds on the top and 3 round on the bottom. Furthermore, we utilize all of the sufficient bit-conditions of the 30-round differential to compute a set of corresponding subkeys. Then we distribute the plaintext pairs over the 2⁸⁶ lists corresponding to the 86-bit subkeys. If a list contains two or more pairs, we regard the subkeys corresponding to the list as candidate subkeys. The time complexity of our attack on 30-round SIMON64/96 (SIMON64/128) is 2^86.2 (2^118.2) with a success probability of 0.61, while the data complexity and the memory complexity are 2^63.3 and 2⁹⁰ bytes, respectively.     

Linear-Differential Cryptanalysis for SPN Cipher Structure and AES

0 Introduction Substitution and permutation network (SPN) structure is one of the most widely used structures in block ciphers. The SPN structure is based on Shannon’s principles of confusion and diffusion[1] and these principles are implemented through …     

11-Round impossible differential algebraic attack for serpent encryption algorithm

This paper explored algebraic features of nonlinear parts in Serpent encryption algorithm and offered an 11-round Serpent-128 impossible differential algebraic attack through utilizing the method in constructing S-box algebraic equations. The new method analyzed block 11-round Serpent with 2¹²⁷ selected plaintexts and 2¹²⁹ bytes memory space at the same time of giving a new design principle of S-box anti-algebraic attack.     

A New Method for Impossible Differential Cryptanalysis of 8-Round Advanced Encryption Standard

This paper first presents an impossible differential property for 5-round Advanced Encryption Standard （AES） with high probability. Based on the property and the impossible differential cryptanalytic method for the 5-round AES, a new method is proposed for cryptanalyzing the 8-round AES-192 and AES-256. This attack on the reduced 8-round AES-192 demands 2^121 words of memory, and performs 2^148 8-round AES-192 encryptions. This attack on the reduced 8-round AES-256 demands 2^153 words of memory, and performs 2^180 8-round AES-256 encryptions. Furthermore, both AES-192 and AES-256 require about 2^98 chosen plaintexts for this attack, and have the same probability that is only 2^-3 to fail to recover the secret key.     

A related-key boomerang distinguishing attack of Threefish-256

The block cipher Threefish is the main component of Skein, which is based on ARX. Based on the efficient algorithms for calculating the differential of modular addition, we extend local collisions of Threefish-256 to more round by using related-key differential of addition in this paper. A related-key boomerang distinguish attack is proposed on 31-round Threefish-256 with a time complexity of 2²³⁴.     

22-轮SMS4的差分分析

SMS4是中国官方公布的第一个商用分组密码标准,使用差分方法分析了18轮的SMS4差分特征,并在此基础上攻击了22-轮的SMS4,攻击过程需要2117个选择明文,2112字节的存储空间,而时间复杂度为2123次22-轮加密。此结果是目前对SMS4差分分析的最好结果。     

Sandwich-Boomerang attack on reduced round CLEFIA

CLEFIA （ named after the French word ＂Clef＂ meaning ＂Key＂ ） is an efficient, highly secure block cipher proposed by SONY Corporation in the 14th International Workshop on Fast Software En- cryption （ FSE-2007 ） and many cryptanalyses have been used to analyze it. According to the proper- ty of CLEFIA, a new technique Sandwich-Boomerang cryptanalysis is used on it. An 8-round Sand- wich-Boomerang distinguisher of CLEFIA is constructed using the best differential characteristic of CLEFIA. And then, based on the distinguisher, an attack against 10-round CLEFIA is proposed. The number of chosen plaintexts required is 2^119（or 2^120） and the time complexity is 2^120（or 2^121）. Compared with a 7-round impossible Boomerang distinguisher presented by Choy in the 4th Interna- tional Workshop on Security （IWSEC-2009） , the differential characteristics used in the attack are all the best ones, so it is believed that the attack is the best result that the Boomerang attacks can get on CLEFIA at present.     

ARIA的不可能差分分析

研究了ARIA在不可能差分分析下的安全性.通过对ARIA线性扩散层的分析,提出一类新的6轮ARIA的不可能差分,并从差分重量的角度,给出了2类具有一般形式的6轮ARIAR的不可能差分的结构和计数,从理论上证明了能够达到目前研究最优的6轮ARIA的所有不可能差分.研究结果表明,在输入输出差分重量为10的条件下,攻击6轮ARIA所需的数据量为2120个选择明文,计算量为294.5次6轮加密.
     

“与常数模2~n加”运算的不可能差分性质研究

对"与常数模2n加"即Y=(X+K)mod2n的不可能差分性质进行了研究,首次给出了当差分转移概率取值为0时,输入差ΔX、输出差ΔY及常数K的结构特点和计数下界。