TY - GEN
T1 - Studying Special Operators for the Application of Evolutionary Algorithms in the Seek of Optimal Boolean Functions for Cryptography
AU - Mandujano, Sara
AU - Ku Cauich, Juan Carlos
AU - Lara, Adriana
N1 - Publisher Copyright:
© 2022, The Author(s), under exclusive license to Springer Nature Switzerland AG.
PY - 2022
Y1 - 2022
N2 - The role of Boolean functions in modern cryptography has triggered the necessity of developing methods to construct them with adequate properties, such as balancedness and high non-linearity—making them more resistant to a variety of cryptanalytic attacks. Research into the construction of weight-wise perfectly balanced Boolean functions using Evolutionary Algorithms is scarce but encouraging (e.g., [1]). In this work, we first investigate the effect on an evolutionary algorithm’s performance when relying solely on the penalty function, as opposed to the solution repairment method. Second, we focus on the effect of problem-specific crossover operators (e.g., those used on [2]), and particularly proposing a novel one free of solution repairs to preserve balancedness. The results obtained suggest that an adequate penalty factor and the use of specifically designed evolutionary operators is sufficient to find Boolean functions with weight-wise perfect balancedness and high non-linearity, as desired.
AB - The role of Boolean functions in modern cryptography has triggered the necessity of developing methods to construct them with adequate properties, such as balancedness and high non-linearity—making them more resistant to a variety of cryptanalytic attacks. Research into the construction of weight-wise perfectly balanced Boolean functions using Evolutionary Algorithms is scarce but encouraging (e.g., [1]). In this work, we first investigate the effect on an evolutionary algorithm’s performance when relying solely on the penalty function, as opposed to the solution repairment method. Second, we focus on the effect of problem-specific crossover operators (e.g., those used on [2]), and particularly proposing a novel one free of solution repairs to preserve balancedness. The results obtained suggest that an adequate penalty factor and the use of specifically designed evolutionary operators is sufficient to find Boolean functions with weight-wise perfect balancedness and high non-linearity, as desired.
KW - Boolean functions
KW - Cryptography
KW - Genetic algorithms
UR - http://www.scopus.com/inward/record.url?scp=85142608346&partnerID=8YFLogxK
U2 - 10.1007/978-3-031-19493-1_30
DO - 10.1007/978-3-031-19493-1_30
M3 - Contribución a la conferencia
AN - SCOPUS:85142608346
SN - 9783031194924
T3 - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
SP - 383
EP - 396
BT - Advances in Computational Intelligence - 21st Mexican International Conference on Artificial Intelligence, MICAI 2022, Proceedings
A2 - Pichardo Lagunas, Obdulia
A2 - Martínez Seis, Bella
A2 - Martínez-Miranda, Juan
PB - Springer Science and Business Media Deutschland GmbH
T2 - 21st Mexican International Conference on Artificial Intelligence, MICAI 2022
Y2 - 24 October 2022 through 29 October 2022
ER -