Abstract:The stability (or instability) of synchronization is important in a number of real world systems, including the power grid, the human brain and biological cells. For identical synchronization, the synchronizability of a network, which can be measured by the range of coupling strength that admits stable synchronization, can be optimized for a given number of nodes and links. Depending on the geometric degeneracy of the Laplacian eigenvectors, optimal networks can be classified into different sensitivity levels, which we define as a network's sensitivity index. We introduce an efficient and explicit way to construct optimal networks of arbitrary size over a wide range of sensitivity and link densities. Using coupled chaotic oscillators, we study synchronization dynamics on optimal networks, showing that cospectral optimal networks can have drastically different speed of synchronization. Such difference in dynamical stability is found to be closely related to the different structural sensitivity of these networks: generally, networks with high sensitivity index are slower to synchronize, and, surprisingly, may not synchronize at all, despite being theoretically stable under linear stability analysis.

What problem does this paper attempt to address?

The problems that this paper attempts to solve mainly focus on the optimization of synchronous networks and their dynamic characteristics. Specifically, the researchers are concerned with how to construct optimal networks that can maximize the range of coupling strengths for stable synchronization given the number of nodes and links. In addition, the paper also explores the differences in synchronization speed among optimal networks with different sensitivity levels and the relationship between these differences and the sensitivity of network structures. ### Main Problems 1. **Construction of Optimal Networks**: - How to construct optimal networks of any size and with different sensitivity levels? - What are the structural characteristics of these optimal networks that enable them to achieve stable synchronization under specific conditions? 2. **Synchronization Stability**: - How is the synchronization stability of a network affected by its structural characteristics? - What are the differences in synchronization speed and time among optimal networks with different sensitivity levels? 3. **Introduction of Sensitivity Index**: - How to quantify the structural sensitivity of optimal networks? - What is the impact of the sensitivity index on the network synchronization performance? ### Specific Research Contents - **Definition of Optimal Networks**: An optimal network refers to a network that can maximize the range of stable - synchronization coupling strengths given the number of nodes and links. - **Sensitivity Index**: The sensitivity index of a network is defined by the size of the largest Jordan block of the network Laplacian matrix. The higher the sensitivity index, the more sensitive the network is to structural perturbations. - **Generation Method**: Using the generalized Krapivsky - Redner network growth model, by controlling the redirection parameter \( r \), optimal networks with different sensitivity indices can be generated. - **Study of Synchronization Dynamics**: By coupling chaotic oscillator systems, the performance of optimal networks with different sensitivity levels in terms of synchronization speed and time is studied. ### Main Findings - **Relationship between Synchronization Speed and Sensitivity**: Generally, networks with a higher sensitivity index have a slower synchronization speed and sometimes may even fail to synchronize, although linear stability analysis predicts that they should be stable. - **Non - global Synchronization Basins**: Highly sensitive networks may "kick" the system state out of the synchronization basin due to the too - rapid growth of initial transient perturbations, resulting in failure to synchronize. ### Conclusion This research provides a method for quantifying the structural sensitivity of optimal networks by introducing the sensitivity index and experimentally verifies the impact of the sensitivity index on synchronization performance. These findings are of great significance for designing and optimizing synchronous networks in practical applications, especially in fields such as power systems and biological systems.

Construction,sensitivity index, and synchronization speed of optimal networks

Optimal Synchronizability of Networks

Network synchronization with periodic coupling

Synchronization on effective networks

SYNCHRONIZATION OF HETEROGENEOUS OSCILLATORS UNDER NETWORK MODIFICATIONS: PERTURBATION AND OPTIMIZATION OF THE SYNCHRONY ALIGNMENT FUNCTION

The essential synchronization backbone problem

Synchronization in Time-Varying Networks.

Structural Effects on Synchronizability of Scale-Free Networks

The Optimum of Synchronizability in Complex Networks

The Efficiency of Synchronization Dynamics and the Role of Network Syncreactivity

Increasing the synchronization stability in complex networks

Effects of dynamical and structural modifications on synchronization.

Network synchronization landscape reveals compensatory structures, quantization, and the positive effect of negative interactions

Networks' modulation: How different structural network properties affect the global synchronization of coupled Kuramoto oscillators

Complex network synchronizability: analysis and control.

Toward Synchronization in Networks With Nonlinear Dynamics: A Submodular Optimization Framework

Optimizing synchronization stability of the Kuramoto model in complex networks and power grids

Network Structure, Topology and Dynamics in Generalized Models of Synchronization

Optimal network design for synchronization of coupled oscillators

Synchronization of Interconnected Heterogeneous Networks: the Role of Network Sizes

Optimal and Poor Synchronizations of Directionally Coupled Phase-Coherent Chaotic Oscillators