Network Strengthening Against Malicious Attacks

Robustness measures the toleration of complex networks against random failures and malicious attacks. A malicious attack removes the most important node iteratively and destroys the network quickly. It is crucial to strengthen the network robustness against malicious attacks. In this paper, we propose an algorithm to strengthen the robustness with reduced change of network structure compared to state-of-the-art algorithms. The algorithm is called targeted variable neighborhood search (TVNS) algorithm. Experiments on real-world and random networks show that TVNS is efficient on strengthening the robustness against malicious attacks. The strengthened network against high degree adaptive attack shows an onion-like structure where nodes prefer to connect with similar degree nodes; while for the strengthened network against high betweenness adaptive attack, nodes prefer to connect with similar betweenness nodes.