Design And Implementation Of Block Cipher Algorithm For Internet Of Things

Powered by the Internet of Things(Io T),the use of resource-constrained Io T smart devices such as sensors and other embedded devices is becoming more widespread,but this increases the risk of their information security.Block ciphers are widely adopted to protect information security by encrypting sensitive data because of their high security,high efficiency and easy hardware and software implementation,especially the application of advanced encryption standard(AES)is an important research hotspot at present.In order to improve the encryption and decryption speed of the block cipher algorithm AES,Intel proposed the AESNI(Intel? Advanced Encryption Standard(AES)New Instructions Set).Since the introduction of the AESNI instruction set,the design of high-speed cryptographic primitives based on the AESNI instruction set has attracted much attention in the industry,such as the ROCCA cipher algorithm for 6G systems.However,in the resource-constrained Io T environment,not all processors support the AESNI instruction set(e.g.ARM Cortex?-M4 microprocessor).How to design cryptographic algorithms based on a high-speed implementation of AES for multiple processors is one of the current research challenges.In this thesis,the optimal implementation techniques of the AES block cipher algorithm and the design methods of existing cipher algorithms with efficient implementations thoroughly researched.Based on which an efficient and secure encryption scheme for Io T is proposed,and the main research results are as follows:(1)Based on the AESNI instruction set and the principle of fixslicing AES implementation,an efficiently implemented encryption algorithm for multiple processors,MPECA(Multi-Platforms Efficient Cryptographic Algorithm),has been designed.This algorithm is an authenticated cipher algorithm based on the block cipher algorithm AES,the round function of MPECA uses four round functions of AES to enable efficient software implementation using fixslicing technology in resource-constrained Io T environments that do not support the AESNI instruction set.In particular,the number of the round functions in the initialization process and the label generation process of MPECA is designed to be 4.MPECA spends 32 fewer AES round function operations and128 fewer XOR operations in the initialization and tag generation processes than ROCCA,which makes its implementation faster in the environments that support AESNI instructions.The experimental results show that compared with ROCCA,the average encryption speed of MPECA is 3.05 times faster in the environment that does not support AESNI instructions,while the average encryption speed of MPECA is 30.64% faster in the environment that supports AESNI instructions.In the ARM implementation environment,the average encryption speed of MPECA is nearly 2.37 times faster than ROCCA.(2)A security analysis of the MPECA cipher algorithm is performed to prove that the MPECA algorithm has sufficient security.To ensure that the MPECA algorithm has good diffusion and obfuscation,the randomness of 100,000 512-bit MPECA cipher sequences is tested by the NIST standard.Furthermore,the byte-oriented mixed integer linear programming(MILP)automated search model of the MPECA algorithm is constructed and the lower bound of the active S-box of the initialization process in MPECA is analyzed.Meanwhile,the linear attacks,forgery attacks,integral attacks were also analyzed.The experimental and analytical results show that the MPECA cipher algorithm has good randomness,and its P-value of frequency test reaches 0.69.MPECA has the ability to resist differential attacks,and the security redundancy of MPECA at round 4 is higher than that of ROCCA at round 20.And MPECA can effectively resist linear attacks,forgery attacks,integral attacks,guess-and-determine attacks,and key recovery attacks after known ciphertext state recovery attack.(3)A SAFE cryptographic algorithm based on a semi-fixslicing technique was designed to solve the issues of ROCCA cannot resist state recovery attacks on known ciphertexts and the high code resources required for fixslicing AES implementation in resource-constrained environments.Specifically,the round function adopts AES quadratic iteration components to enable efficient implementation of SAFE in resource-constrained embedded devices based on semi-fixslicing.The number of ROCCA’s round functions was reduced from 20 to 9 in the initialization and label generation processes based on differential analysis.To improve the security of SAFE algorithm,a new nonlinear component was introduced in its encryption process.The experimental results show that the SAFE algorithm has high security,and can better resist state recovery attacks on known ciphertexts than the ROCCA algorithm.In ARM Cortex series embedded devices,the SAFE algorithm improves the implementation speed of data encryption and decryption,and its encryption and decryption speed is 2.06 times faster than the ROCCA algorithm.