Analyzing Complex Non-Trivial Network using Attack Set Generation by Genetic Algorithm

Nowadays, security of the networks is one of the major concern. Attack paths in an attack graph give a way to get a view of the big network, illustrating all the possible vulnerabilities in a network, from a security point of view. This paper proposes a new methodology for finding all the possible attack paths in a graph. It helps us in identifying most desirable and least desirable attack paths by the attacker, which will give network administrators a view for securing their network. Some researchers have used a genetic algorithm (GA) for finding the attack paths as GA helps us in providing a fast way to generate the possible list of solutions in very less time. We have also used this genetic algorithm but in a different, better and modified way for our approach by introducing a new scheme of backward mutation, with 100 percent GA operators(crossover, mutation) rate and by also modifying the phases of GA for generating fast results. By performing experiments, our new modified approach for GA is producing 7 percent (approx.) more solutions by keeping same parameters as that of existing GA. Other algorithms may also tell us about all attack paths but they will either be slow or may miss out some attack paths in a network.