SkyRoad network: A hierarchical graph-based framework for safe and efficient low-altitude airspace utilization

Urban low-altitude airspace represents an emerging three-dimensional resource for alleviating ground traffic congestion, yet its strong spatial heterogeneity, pronounced spatiotemporal dynamics, and diverse operational entities pose substantial challenges for structured utilization and management. This study proposes the SkyRoad Network (SkyRN), a hierarchical framework that transforms continuous low-altitude airspace into an attributed route network through multi-level H3 hexagonal grid discretization and adaptive spatial merging. Multi-source geographic data, including building distribution, population density, and infrastructure layout, are integrated at the grid-cell level, yielding a multidimensional attribute model that captures time-varying exposure risk, multi-buffer-zone accessibility, and a discrete operational-status code carried at the edge level. A dual-attribute-driven adaptive merging strategy reduces computational complexity while preserving safety-critical spatial information, yielding smoother path geometry with reduced turning frequency on the merged network. By maintaining the operational status as a dynamic edge attribute over an otherwise invariant topology, a single network supports both offline route planning and real-time replanning under emergent disruptions, and performance-differentiated routing is obtained by weighting the shared edge attributes. A case study in Shenzhen validates the framework’s applicability under diverse drone performance configurations, demonstrating configuration-dependent altitude and route utilization, and real-time adaptation to localized disruptions through per-edge status updates. Structural comparison with the ground transportation network shows that the multi-layer low-altitude network attains substantially higher resilience through vertical redundancy, with robustness distributed across altitude layers. By coupling city-scale structural simplification with an edge-level dynamic state, SkyRN offers a practical foundation for the joint operation and supervision of safe and efficient low-altitude airspace utilization.