Security Analysis Of Block Cipher Based On Automated Search

With the widespread use of Internet of Things technology and smart phones today,the Internet has completely penetrated every corner of the world.At the same time,the issue of network security has become increasingly prominent.Cryptography,the basis of network security,plays an irreplaceable role in maintaining user information security.Block ciphers are always the hot point of cryptographers research due to its high efficiency and suitability for large-scale data encryption.With the rapid development of block cipher design,various block cipher analysis techniques are also becoming mature.The quality of cryptanalysis technology directly affects the evaluation of the security degree of block cipher algorithms.With the development of automated cryptanalysis,it shows that the automated search in cryptanalysis greatly simplify the tedious manual derivation,and can obtain better results in a short time.This paper mainly uses the bit-based integral division property proposed by Japanese scholar Todo in 2015 to study cryptographic algorithms,and combined with automated search technology to study cryptographic algorithms.The integral division property is divided more explicitly than traditional integral analysis does.Therefore searching the integral distinguisher of the algorithm is more effectively,especially for the bit-based block encryption algorithm.In this paper,we use different automatic search methods for the different structure of the cipher algorithm,and study the security degree of four cipher algorithms based on the analysis of bit division property.They can be divided into integral analysis of SPN structure,Feistel structure based on MILP method and integral analysis of ARX structure based on SAT method.The specific research content is as follows.1.A variety of different integral distinguishers were found by automatically searching the integral distinguisher of SPN structure algorithm GIFT-64 with the MILP method,which includes the longest 9-round integral distinguisher so far.And an ordinary round was extended in the direction of decryption.At the same time,13-round and 14-round key recovery attacks were discovered on the basis of 10-round distinguishers,and the complexity was analyzed respectively.2.A variety of different integral distinguishers were found by automatically searching the integral distinguisher of Feistel structure algorithm GRANALE with the MILP method,which includes the longest 10-round integral distinguisher so far.At the same time,13-round key recovery attacks were discovered on the basis of 9-round distinguishers,and the complexity was analyzed respectively.3.A variety of different integral distinguishers were found by automatically searching the integral distinguisher of Feistel structure algorithm MANTRA with the MILP method,which includes the longest 11-round integral distinguisher so far.At the same time,14-round key recovery attacks were discovered on MANTRA-64 with a key length of 64 bits,and 16-round key recovery attacks were discovered on MANTRA-128 with a key length of 64 bits.And the complexity was analyzed respectively.4.A variety of different integral distinguishers were found by automatically searching the integral distinguisher of an ARX-type generalized four-branch Feistel structure algorithm CHAM with the SAT method.For CHAM-64 with a block length of 64 bits,a variety of integral distinguishers were found,including the longest 24-round integral distinguisher.For the CHAM-128 with a block length of 128 bits,a variety of integral distinguishers are also found,including the longest 24-round integral distinguisher.At the end of this paper,we summarize the whole paper,expound the difficulties of automatic cryptanalysis technology,and discuss the development and Prospect of automatic cryptanalysis technology.