A Novel S-box Optimization Method Based on Immune Genetic Algorithm

Ding Zhu,Miao Zhang,Xiaojun Tong,Zhu Wang

DOI: https://doi.org/10.1109/icsess49938.2020.9237665

2020-01-01

Abstract:S-box is the only nonlinear component in block cipher, and its performance is the key factor to determine the data security of cipher algorithm. Our paper proposes a novel S-box optimization method based on immune genetic algorithm. Firstly, the S-boxes population is generated by chaotic system, then the S-boxes with excellent performance are optimized by a series of operators, including extracting the anti-agent and immune selection. The nonlinear degree criterion, differential uniformity criterion and strict avalanche effect criterion of S-boxes are analyzed. The experimental results show that the optimized S-box has strong characteristics. Compared with the traditional genetic algorithm, this method has faster convergence speed and better resistance to linear attacks and differential attacks. The optimized S-box has a good application in the field of information security prospects.