On evolutionary computation for moving target defense in software defined networks
Adetokunbo Makanju, Nur Zincir-Heywood, Shinsaku Kiyomoto · Proceedings of the Genetic and Evolutionary Computation Conference Companion · 2017
Moving Target Defense (MTD) is a paradigm in cyber-security that proposes cyber-systems that continually change their configurations to avoid easy surveillance from adversaries. Despite its promise, MTDs have yet to be fully adopted in real world systems. Using the concepts from the general theory of MTD, this paper proposes that the goal of MTDs i.e continually adapting to their environment to evade attack is similar to the concept of an organism evolving to survive in its habitat. This implies that Evolutionary Computation (EC) techniques could be used to fulfill the goal of MTD, specifically in relation to intrusion detection in the emerging field of Software Defined Networks (SDNs). The programmable nature of SDNs provide a degree of flexibility not possible in conventional networks. Thus, making it possible to completely automate the configuration of such networks.