Solving Linear Equations Systems Using Genetic Algorithm

Abstract

Genetic Algorithms (GA's) are a class of optimization algorithms. GA’s attempts to solve problems through modeling a simplified version of genetic process. There are many problems for which a GA approach is useful.
This paper aims to solve Linear Equations System (LES) for any number of variables using the GA. The application of this paper represented by cryptanalysis application, this done by attacking stream cipher systems, choosing one Linear Feedback Shift Register (LFSR), since its considered as a basic unit of stream cipher systems, in the performance of GA. The application divided into two stages, first, constructing LES’s for the LFSR, and the second, is attacking the variables of LES’s which they are also the initial key values the of LFSR.