Malware Diffusion Models for Modern Complex Networks

5.3. Malware Diffusion Modeling Based on MRFs

One of the key observations regarding the dynamics of malware diffusion in various types of networks, and especially highlighted in wireless ad hoc networks as shown in the previous chapter, is dictated by interactions of adjacent nodes (corresponding to users directly connected with each other). Such local interactions can be interpreted as local dependencies of the corresponding users/nodes and this triggers employing MRFs for modeling such interactions/dependencies. Since MRFs generalize the notion of the spatial Markov property [136], it will become possible to capture the aforementioned dependencies in a holistic manner, uniformly for all types of complex communications networks. In this section, ...