1: \begin{abstract}

2:   This paper studies network resilience against structured additive

3:   perturbations to its topology. We consider dynamic networks modeled

4:   as linear time-invariant systems subject to perturbations of bounded

5:   energy satisfying specific sparsity and entry-wise constraints.

6:   Given an energy level, the structured pseudospectral abscissa

7:   captures the worst-possible perturbation an adversary could employ

8:   to de-stabilize the network, and the structured stability radius is

9:   the maximum energy in the structured perturbation that the network

10:   can withstand without becoming unstable.  Building on a novel

11:   characterization of the worst-case structured perturbation, we

12:   propose iterative algorithms that efficiently compute the structured

13:   pseudospectral abscissa and structured stability radius.  We provide

14:   theoretical guarantees of the local convergence of the algorithms

15:   and illustrate their efficacy and accuracy on several network

16:   examples.

17: \end{abstract}

18: