Agent-based simulation of the dynamics of malware propagation in scale-free networks

Abstract

Agent-based modeling and simulation (ABMS) has been used in wide range of application areas as an effective modeling paradigm. In this work, we have used ABMS to study the dynamics of malware propagation in scale-free networks (SFNs). Firstly, we use an analytical model to formulate the malware propagation process that considers the diversity of nodes in SFNs. Secondly, for the same problem, we use agent-based simulation to model the outbreak of malware based on the rumor diffusion process. The proposed agent-based model considers a set of heterogeneous agents and interactions between them that may allow malware transmission. This model helps us to study the effects of defense mechanisms, such as software diversity and immunization, in controlling the malware propagation and thereby reducing the impacts of attacks in networks. We also investigate the effect of the vulnerability function related to the degree of network nodes to explain the diversity of node anti-attack ability in the agent-based model. The results of the agent-based simulation are nearly the same as the analytical model, which can be used to validate the simulation model. However, the agent-based model has some notable advantages, such as flexibility and dynamic reconfiguration, over the analytical model.