A survey of multilayer networks modelled to assess robustness in infrastructure systems

Abstract

The development of modern societies places particular demands on the consistent performance of infrastructure systems. Because multilayer network models are capable of representing the interdependencies between infrastructure components, they have been widely used to analyse the robustness of infrastructure systems. This study is a systematic review of the literature published since 2010. It aims to investigate how multilayer network models have been used in analysing the robustness of infrastructure systems. According to the findings, the percolation theory was the most popular method, used in about 57% of papers. Regarding the properties, coupling strength and node degree were the most common, while directed links and feedback conditions were the least common. The following gaps were identified, which provide opportunities for further research. These include the absence of models based on real-world data and the need for models that make fewer simplifying assumptions about complex systems. No papers considered all potential properties and their effect on boosting or weakening each other’s effect. By considering all properties, the importance of different properties on the robustness of infrastructure systems can be quantified and compared in future studies.