Characterization of Driver Nodes of Anti-Stable Networks
For researchers in network control theory, this work addresses the challenge of selecting optimal driver nodes in anti-stable networks, though it is limited to networks controllable with a single driver node.
This work characterizes driver nodes for anti-stable networks and provides theory and algorithms to select driver nodes that maximize the region of attraction under limited capacity, addressing the nonuniqueness of driver nodes.
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.