Interplay Between Network Heterogeneity and Activation Thresholds in Global Cascades

Abstract

The onset of global cascades in complex networks depends on both the distribution of activation thresholds and the underlying network topology. Here we investigate how these two sources of heterogeneity interact within the Watts threshold model, using beta-distributed activation thresholds and beta-binomial degree distributions to systematically vary their respective properties. Analyzing the onset of cascades from infinitesimal seeds, we identify two distinct boundaries governing the large-connectivity behavior of vulnerable branching. The first separates regimes in which increasing connectivity enhances or suppresses propagation, depending on the abundance of individuals with low activation thresholds. The second distinguishes two mechanisms of suppression: one controlled by the scarcity of low-threshold individuals and the other by the diminishing abundance of low-degree nodes as connectivity increases. At the boundary between these mechanisms, their competing contributions produce a logarithmic correction to the asymptotic scaling. Our results demonstrate how the interplay between network structure and individual activation thresholds determines the conditions for extensive propagation.