System Relaxation for Interpretable and Adaptive Network Control
Fuente:
arXiv
Guardado en:
| Autores principales: | , , |
|---|---|
| Formato: | Preprint |
| Publicado: |
2025
|
| Materias: | |
| Acceso en línea: | |
| Etiquetas: |
Agregar Etiqueta
Sin Etiquetas, Sea el primero en etiquetar este registro!
|
| _version_ | 1866911166304878592 |
|---|---|
| author | Ren, Zhiyuan Shuai, Zhiliang Cheng, Wenchi |
| author_facet | Ren, Zhiyuan Shuai, Zhiliang Cheng, Wenchi |
| contents | Prevailing network control strategies, which rely on static shortest-path logic, suffer from catastrophic "stress concentration" on critical nodes. This paper introduces the System Relaxation Algorithm (SRA), a new control paradigm inspired by physical relaxation that guides a network toward an emergent equilibrium of load balance. SRA is an interpretable, 'white-box' dynamical system whose behavior is profoundly topology-dependent: in heterogeneous networks, it acts as a proactive performance optimizer, reducing peak centrality by over 80\% and increasing high-load throughput by more than 45\%; in homogeneous topologies, its objective intelligently shifts to resilience enhancement. We rigorously prove its global convergence and practical stability using the theory of non-smooth dynamical systems, establishing a predictable paradigm for network governance that intelligently trades off performance and resilience. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2509_16984 |
| institution | arXiv |
| publishDate | 2025 |
| record_format | arxiv |
| spellingShingle | System Relaxation for Interpretable and Adaptive Network Control Ren, Zhiyuan Shuai, Zhiliang Cheng, Wenchi Networking and Internet Architecture Systems and Control Prevailing network control strategies, which rely on static shortest-path logic, suffer from catastrophic "stress concentration" on critical nodes. This paper introduces the System Relaxation Algorithm (SRA), a new control paradigm inspired by physical relaxation that guides a network toward an emergent equilibrium of load balance. SRA is an interpretable, 'white-box' dynamical system whose behavior is profoundly topology-dependent: in heterogeneous networks, it acts as a proactive performance optimizer, reducing peak centrality by over 80\% and increasing high-load throughput by more than 45\%; in homogeneous topologies, its objective intelligently shifts to resilience enhancement. We rigorously prove its global convergence and practical stability using the theory of non-smooth dynamical systems, establishing a predictable paradigm for network governance that intelligently trades off performance and resilience. |
| title | System Relaxation for Interpretable and Adaptive Network Control |
| topic | Networking and Internet Architecture Systems and Control |
| url | https://arxiv.org/abs/2509.16984 |