2022/02/01 by Jeremie Fish, Erik M. Bollt, Fish, Jeremie +1
Biochemistry, Genetics and Molecular Biology · Computer Science · #Chaotic Dynamics (nlin.CD) #Dynamical Systems (math.DS) #FOS: Mathematics #FOS: Physical sciences #Gene Regulatory Network Analysis #Neural Networks Stability and Synchronization #Nonlinear Dynamics and Pattern Formation
paper · pdf · doi:10.48550/arxiv.2202.00156
openalex publication_date 2022/02/01 · openalex created_date 2025/10/10 · openalex updated_date 2026/07/28
It has been recognized for quite some time that for some matrices the spectra are not enough to tell the complete story of the dynamics of the system, even for linear ODEs. While it is true that the eigenvalues control the asymptotic behavior of the system, if the matrix representing the system is non-normal, short term transients may appear in the linear system. Recently it has been recognized that since the matrices representing directed networks are non-normal, analysis based on spectra alone may be misleading. Both a normal and a non-normal system may be stable according to the master stability paradigm, but the non-normal system may have an arbitrarily small attraction basin to the synchronous state whereas an equivalent normal system may have a significantly larger sync basin. This points to the need to study synchronization in non-normal networks more closely. In this work, various tools will be utilized to examine synchronization in directed networks, including pseudospectra, an adaption of pseudospectra that we will call Laplacian pseudospectra. We define a resulting concept that we call Laplacian pseudospectral resilience (LPR). It will be shown that LPR outperforms other scalar measures for estimating the stability of the synchronous state to finite perturbations in a class of networks known as optimal networks. Finally we find that the ideal choice of optimal network, with an eye toward synchronization, is the one which minimizes LPR