Genetic Algorithm Stream Cipher Key Generation Using NIST Functions

ECU Author/Contributor (non-ECU co-authors, if there are any, appear on document)

Hunter Wayne Leggett (Creator)

Institution

East Carolina University (ECU )

Web Site: http://www.ecu.edu/lib/

Abstract: Stream ciphers are beneficial because of their efficiency, speed, and low resource utilization. However, stream ciphers are vulnerable to many attacks if they do not use strong keys for encryption and decryption. Thus, one way to increase the security of stream ciphers is to improve the key generation algorithm. This study sought to evaluate the keys produced by a genetic algorithm stream cipher when individual and combinations of fitness functions are used. Furthermore, this study identified which fitness function is the best for a specific scenario. For the genetic algorithm, the fitness tests are thirteen of the tests defined in NIST SP 800-22rla. The thirteen different fitness functions were inputted into the genetic algorithm stream cipher one at a time. Next, 50 total keys of varying bit sizes were generated. These keys were evaluated by using the Hamming distance between the keys and time that it took for key generation. After each individual fitness function was evaluated, two combinations of five tests were created and used as a single fitness function. The two combinations were the best performing NIST functions for Hamming distance and time for 256-bit keys. Sensitivity analysis was then performed to find the best possible combination of the NIST functions. Based on the results, using different individual functions or a combination of functions as a fitness functions changed the Hamming distance between the keys and the time that it takes to generate a key. Furthermore, using sensitivity analysis results for the top two, three, four, and five combinations for Hamming distance and time, prediction equations were created and used to predict values for other combinations and key sizes.

Additional Information

Publication

Thesis

Language: English

Date: 2023

Subjects

Cryptography;genetic algorithm;Hamming distance;cybersecurity;stream cipher

Email this document to