A relaxed small-gain theorem for infinite networks

TL;DR

This paper introduces a relaxed small-gain theorem for infinite networks, enabling stability analysis of large interconnected systems and transferring results from infinite to finite networks without knowing their size.

Contribution

It develops a novel small-gain theorem for infinite networks, extending stability analysis methods and allowing finite network stability inference without size knowledge.

Findings

The theorem guarantees stability transfer from infinite to finite networks.

It applies to urban traffic networks, demonstrating practical effectiveness.

The approach does not require the size of the finite network.

Abstract

Motivated by the scalability problem in large networks, we study stability of a network of infinitely many finite-dimensional subsystems. We develop a so-called relaxed small-gain theorem for input-to-state stability (ISS) with respect to a closed set and show that every exponentially input-to-state stable system necessarily satisfies the proposed small-gain condition. Following our bottom-up approach, we study the well-posedness of the interconnection based on the behavior of the individual subsystems. Finally, we over-approximate large-but-finite networks by infinite networks and show that all the stability properties and the performance indices obtained for the infinite system can be transferred to the original finite one if each subsystem of the infinite network is individually ISS. Interestingly, the size of the truncated network does not need to be known. The effectiveness of our small-gain theorem is verified by application to an urban traffic network.