Multiple scale theory of topology driven pattern on directed networks

TL;DR

This paper develops a multiple scale theory for pattern formation on directed networks driven by topology, deriving an effective Stuart-Landau equation to describe nonlinear amplitude evolution near instability thresholds.

Contribution

It introduces a novel multiple scale perturbative approach to derive a Stuart-Landau equation for reaction-diffusion systems on directed graphs, linking topology to nonlinear pattern dynamics.

Findings

Theory matches well with numerical simulations.

Derived amplitude equation depends on network topology.

Identified conditions for pattern formation on directed networks.

Abstract

Dynamical processes on networks are currently being considered in different domains of cross-disciplinary interest. Reaction-diffusion systems hosted on directed graphs are in particular relevant for their widespread applications, from neuroscience, to computer networks and traffic systems. Due to the peculiar spectrum of the discrete Laplacian operator, homogeneous fixed points can turn unstable, on a directed support, because of the topology of the network, a phenomenon which cannot be induced on undirected graphs. A linear analysis can be performed to single out the conditions that underly the instability. The complete characterization of the patterns, which are eventually attained beyond the linear regime of exponential growth, calls instead for a full non linear treatment. By performing a multiple time scale perturbative calculation, we here derive an effective equation for the non linear evolution of the amplitude of the most unstable mode, close to the threshold of criticality. This is a Stuart-Landau equation whose complex coefficients appear to depend on the topological features of the embedding directed graph. The theory proves adequate versus simulations, as confirmed by operating with a paradigmatic reaction-diffusion model.