Security Attack Detection using Genetic Algorithm (GA) in Policy Based Network

Abstract

We present a modern approach of detecting the computer network security attacks (commonly termed as "hacks") by using an artificial intelligence method known as a genetic algorithm, which is a variant of stochastic beam search, which can be applied to a Policy Based Network. For this experiment, we develop the corresponding form of genetic algorithm and the form of fitness function to detect security attacks. Then we test the algorithm by a simulation program in TC (Turbo C) to gauge its effectiveness and the side effects of its execution. Finally, we give the indication for the necessary change in policy framework to deploy the functionality.