A Powerful Genetic Algorithm to Crack a Transposition Cipher

Abstract

Security of information systems depends heavily on the strength of the cryptosystem. Throughout the years, several cryptosystem algorithms have been developed. These algorithms may inherit some weakness that can jeopardize the integrity of the data. In this paper, we present a genetic algorithm to crack a transposition cipher that extends on the notable research in the field and introduces new ideas including a novel crossover function, a dictionary of the most used words in the English language to evaluate the fitness of the keys in any generation, a dynamic selection method, and a variable generating number. This algorithm starts with a very small randomly selected set of keys, and proceeds the crossover operation on the highly fit keys to produce next generations until a specific number of generations, the final result produces a key that is either a perfect match to the original encryption key or one that is very close. Our experiments and results show mostly optimal solutions for the keys in linear time performance which is a dramatic improvement to the brute force algorithm that takes a factorial time to crack the key.