Node Removal Vulnerability of the Largest Component of a Network

TL;DR

This paper investigates how removing specific nodes affects the largest component of a network, proposing a novel matrix norm minimization approach and a greedy algorithm validated on power grid data.

Contribution

It introduces a new method linking node removal to matrix one-norm minimization and develops a greedy algorithm for targeted node removal.

Findings

The proposed method effectively reduces the largest component size.

The greedy algorithm outperforms traditional centrality-based methods.

Experimental validation on power grid data confirms its efficiency.

Abstract

The connectivity structure of a network can be very sensitive to removal of certain nodes in the network. In this paper, we study the sensitivity of the largest component size to node removals. We prove that minimizing the largest component size is equivalent to solving a matrix one-norm minimization problem whose column vectors are orthogonal and sparse and they form a basis of the null space of the associated graph Laplacian matrix. A greedy node removal algorithm is then proposed based on the matrix one-norm minimization. In comparison with other node centralities such as node degree and betweenness, experimental results on US power grid dataset validate the effectiveness of the proposed approach in terms of reduction of the largest component size with relatively few node removals.