Characterization of Driver Nodes of Anti-Stable Networks

TL;DR

This paper analyzes how to select optimal driver nodes in anti-stable networks to maximize the controllable region, considering capacity limits and eigenvalue instability, providing new algorithms for driver node selection.

Contribution

It introduces novel theory and algorithms for selecting driver nodes in anti-stable networks to maximize the region of attraction, addressing non-uniqueness in driver node selection.

Findings

Algorithms effectively select driver nodes for maximum controllability.

The approach accounts for limited driver node capacity and unstable eigenvalues.

The work demonstrates improved control strategies in anti-stable networks.

Abstract

A controllable network can be driven from any initial state to any desired state using driver nodes. A set of driver nodes to control a network is not unique. It is important to characterize these driver nodes and select the right driver nodes. The work discusses theory and algorithms to select driver node such that largest region of attraction can be obtained considering limited capacity of driver node and with unstable eigenvalues of adjacency matrix. A network which can be controllable using one driver node is considered. Nonuniqueness of driver node poses a challenge to select right driver node when multiple possibilities exist. The work addresses this issue.