Geometry-informed modeling and evaluation of coverage resilience in complex communication networks

The development of 5 G technology drives economic and societal advancement. However, the expanding scale and increasing topological complexity of communication networks pose challenges for modeling higher-order coverage dependencies and assessing network performance. Here, we propose a novel framework for modeling and assessing complex communication networks by integrating computational geometry with hypergraph theory. The system is first represented as a Voronoi-based geometric network (VGN), where pairwise coverage dependencies form the basis graph. We then introduce hyperedges to capture higher-order dependencies among base stations. The resulting hypergraph is subsequently transformed into a line graph, on which we compute a modified augmented Forman–Ricci curvature and project it back to derive a geometry-informed station-level metric. This metric exhibits a strong negative correlation with the loss of coverage (LoC) measures, enabling the evaluation of network resilience. Numerical experiments demonstrate its effectiveness in capturing both first-order (single-station failure) and second-order (joint-neighbor failures) LoC effects. Using a real-world deployment dataset, we further validate the framework’s practical utility for modeling and evaluating complex communication networks. These results establish an efficient, geometry-based approach for quantifying resilience in next-generation communication networks.