An Artificial Arms Race: Could it Improve Mobile Malware Detectors?

Abstract

On the Internet today, mobile malware is one of the most common attack methods. These attacks are usually established via malicious mobile apps. To combat this threat, one technique used is the deployment of mobile malware detectors. As the mobile threats evolve, designing and developing mobile malware detectors remains a challenging task. In this paper, we aim to explore whether creating an artificial arms race between mobile malware and detectors could improve the ability of the detector to adapt to the evolving threats. To better model this interaction, we present a co-evolution of both sides of the arms race using genetic algorithms. The experimental evaluations on publicly available malicious and non-malicious mobile apps and their variants generated by the artificial arms race show that this approach improves the detectors understanding of the problem.