3D City Model

Propagation models use standardized CityGML LoD2 or LoD3 representations. LoD2 provides block models with differentiated roof structures, sufficient for macro-cell path loss. LoD3 adds detailed wall and roof superstructures—parapets, dormers, and balconies—critical for accurate small-cell and mmWave propagation where scattering from fine architectural features dominates the channel impulse response.

A deterministic computational propagation model that simulates the multipath trajectories of radio waves by calculating reflections, diffractions, and scattering from the 3D geometric database of buildings and terrain. Unlike empirical models, ray tracing uses the 3D City Model to identify every potential signal path between transmitter and receiver, accounting for:

• Specular reflections off building facades
• Diffraction around roof edges and corners
• Penetration loss through walls based on material properties The result is a highly accurate, site-specific prediction of path loss, delay spread, and angle of arrival for urban small-cell planning.

A bare-earth 3D raster representation of terrain surface topography that serves as the foundational layer beneath the 3D City Model. The DEM provides the ground height at every grid point, enabling the propagation engine to calculate:

• Line-of-sight obstructions from natural terrain
• Effective antenna heights above ground
• Diffraction loss over hills and ridges When combined with building footprints and heights, the DEM completes the vertical reference for the entire urban propagation environment. Standard resolutions range from 1m to 30m depending on the frequency band and required accuracy.

A critical metadata layer attached to the 3D City Model that defines the complex permittivity, conductivity, and thickness of building surfaces. These electromagnetic material parameters determine:

• Reflection coefficients at each interaction point
• Transmission loss through walls and windows
• Scattering patterns from rough surfaces like brick or concrete Common materials modeled include reinforced concrete, low-emissivity glass, brick, and metal cladding. Without accurate material properties, ray tracing predictions can deviate by 10-20 dB from real-world measurements in dense urban canyons.

A discrete global grid system developed by Uber that partitions the Earth's surface into hierarchical hexagonal cells, providing a standardized spatial indexing system for aggregating and querying 3D City Model data. Key advantages for propagation modeling include:

• Uniform area cells that minimize distortion compared to square grids
• Hierarchical resolution from 0.5m² to 4.3M km²
• Efficient indexing for spatial joins between building data and RF predictions The H3 grid enables rapid aggregation of ray tracing results into coverage maps and facilitates the storage of pre-computed propagation paths for real-time spectrum allocation queries.

A high-fidelity, continuously synchronized virtual replica of a physical electromagnetic environment that integrates the 3D City Model with real-time sensor data and propagation predictions. The digital twin enables network operators to:

• Simulate the impact of new building construction on existing coverage
• Test spectrum sharing policies before deployment
• Optimize small-cell placement using current geometry Unlike a static 3D model, the RF digital twin incorporates dynamic elements such as moving vehicles, temporary structures, and seasonal foliage changes that affect propagation. This concept extends the 3D City Model from a planning tool to an operational decision-support system.