Application of bio-inspired algorithm to the problem of integer factorization

Roman V. Yampolskiy, University of Louisville


This paper describes an attempt at developing an evolutionary algorithm capable of solving non-trivial cases of integer factorisation, which are at the heart of security behind the modern public key cryptography systems. After reviewing previous work in integer factorisation and describing the developed genetic algorithm the paper addresses issues of convergence to a local maxima associated with the performance of genetic algorithms. Specifically, properties and statistical distribution of local maxima points associated with the integer factorisation problems are reviewed. Finally, performance of the developed system is analysed and recommendations are made for future research paths. Copyright © 2010 Inderscience Enterprises Ltd.