Robustness of network controllability to degree-based edge attacks

Abstract

We analyze the tolerance of network controllability to degree-based edge attacks as well as random edge failure. In particular, we leverage both control-based and reachability-based robustness metrics to investigate the case when a fixed number of controls are allowed to change locations following each attack. This ability to change the locations of controls models the more realistic scenario in which operators may have a fixed budget of resources but that these resources can be redeployed in response to attacks on the system. We also identify that the most potent targeted attack for network controllability selects edges (on average) based on betweenness centrality.