Malware propagation in Wireless Sensor Networks: global models vs Individual-based models

Abstract

The main goal of this work is to propose a new framework to design a novel family of mathematical models to simulate malware spreading in wireless sensor networks (WSNs). An analysis of the proposed models in the scientific literature reveals that the great majority are global models based on systems of ordinary differential equations such that they do not consider the individual characteristics of the sensors and their local interactions. This is a major drawback when WSNs are considered. Taking into account the main characteristics of WSNs (elements and topologies of network, life cycle of the nodes, etc.) it is shown that individual-based models are more suitable for this purpose than global ones. The main features of this new type of malware propagation models for WSNs are stated.