Supercriticality of the SIRS on random networks
Abstract
We analyze the supercritical regime of the SIRS process (Susceptible-Infected-Recovered-Susceptible) on random graphs and Galton-Watson trees.
Our results provide a three-part characterization of this process: first, we prove that even on finite sparse graphs, the epidemic can survive for an exponentially long time.
On heavy-tailed networks with power-law-like exponents, we show that this holds for the entire range of parameters.
Notably, while contemporary works establish a stretched exponential survival time for power-law networks, our approach provides the optimal bound of order $e^{\Theta(n)}$ via a different approach.
Finally, for infinite trees, we establish sufficient conditions for strong survival, proving that the root is re-infected infinitely often even for light-tailed degree distributions like the Poisson distribution.