Malware Diffusion Models for Modern Complex Networks

Chapter 5

Malware-propagative Markov random fields

Abstract

In the previous chapter, a queuing theory based framework for modeling malware diffusion was presented and analyzed, capitalizing on analytic tools for closed queuing systems. It allowed modeling the diffusion over both static fixed-topology networks and dynamic networks with churn. Results were presented for both nonpropagative and propagative types of networks yielding a very powerful and generic framework with numerous potentials for network designers and scientists. In this chapter, a different but still probabilistic framework for modeling malware diffusion is presented. This framework differs from the previous as it utilizes models developed in statistical mechanics and more specifically ...