Optimizing communication network geodiversity for disaster resilience through shielding approach

The communication network is the critical infrastructure and backbone of modern society. On the other hand, regional failures, including natural disasters and malicious attacks, pose a major threat to its capacity and quality of service. Therefore, it is important to design a communication network that is resilient enough to withstand disasters. A key factor of network resilience is diversity, as multiple unrelated elements prevent them from sharing the same fate. This paper focuses on the geodiversity of the network and takes it as the resilience metric for the network under geographically highly correlated regional failures. The communication networks with high geodiversity are designed by shielding critical components, where the shielded ones can survive disasters. One of the attractive advantages of the approach is its implementability, more than just a theoretical technique. Meanwhile, we develop an integer nonlinear programming (INLP) model to obtain the minimum cost shielding for communication networks with desired geodiversity. Since the problem is difficult to solve in polynomial time, a genetic-based heuristic algorithm is applied to get a near-optimal solution in a shorter time. Finally, simulation experiments are performed to illustrate our methodology and its effectiveness, and the results highlight the collective role of network elements.