Modeling critical connectivity constraints in random and empirical networks

Abstract

Random networks are a powerful tool in the analytical modeling of complex networks as they allow us to write approximate mathematical models for diverse properties and behaviors of networks. One notable shortcoming of these models is that they are often used to study processes in terms of how they affect the giant connected component of the network, yet they fail to properly account for that component. As an example, this approach is often used to answer questions such as how robust is the network to random damage but fails to capture the structure of the network before any inflicted damage. Here, we introduce a simple conceptual step to account for such connectivity constraints in existing models. We distinguish network neighbors into two types of connections that can lead or not to a component of interest, which we call critical and subcritical degrees. In doing so, we capture important structural features of the network in a system of only one or two equations. In particular cases where the component of interest is surprising under classic random network models, such as sparse connected networks, a single equation can approximate state-of-the art models like message passing which require a number of equations linear in system size. We discuss potential applications of this simple framework for the study of infrastructure networks where connectivity constraints are critical to the function of the system.