What Geospatial Data Layers Are Mandatory vs Optional for a Telecom Project?

For 5G telecom projects, three layers are universally mandatory: a Digital Terrain Model (DTM) for ground elevation, Land Use / Land Cover (LULC) / clutter data for propagation attenuation, and a road or administrative boundary layer for geographic reference. 3D building vectors, vegetation height data, and population density maps become mandatory for urban deployments at 3.5 GHz and above, and for any 5G FWA project. All other layers are optional but add measurable planning accuracy.

What Is the Logic Behind Mandatory vs Optional?

A geospatial data layer is mandatory for a telecom project if omitting it causes the propagation model to produce unreliable results that cannot be compensated by calibration or model adjustment. A layer is optional if its absence degrades accuracy but the model still produces useful planning output.

This distinction shifts depending on:

• Frequency band: higher frequencies interact more precisely with geometry
• Environment type: urban vs. suburban vs. rural
• Project objective: macro coverage vs. FWA qualification vs. small cell placement
• Propagation model type: empirical vs. deterministic

Summary Table: Mandatory vs Optional by Project Type

The Mandatory Layers in Detail

1. Digital Terrain Model (DTM)

A DTM is the foundational geographic layer for any propagation calculation. Without accurate bare-earth elevation, the model cannot compute path elevation profiles, terrain diffraction, or ground reflection geometry.

What to require:

• Bare earth only (buildings and vegetation removed)
• Horizontal resolution: 25 m for rural sub-1 GHz, 5 m for 3.5 GHz urban, 1 to 2 m for 26 GHz
• Vertical RMSE: ≤5 to 10 m (SRTM-grade, rural), ≤2 m (urban 3.5 GHz), ≤1 m (mmWave)
• Hydro-enforced version recommended for areas with rivers, lakes, or coastal terrain

What to avoid:

• Using a DSM (Digital Surface Model) as a DTM proxy: the DSM includes building and tree heights merged into the terrain surface
• SRTM for urban or 3.5 GHz+ planning: 30 m resolution and 5 to 10 m RMSE are too coarse

2. LULC / Clutter Classification

Every propagation model requires a land-use / land-cover layer to assign attenuation coefficients by environment type. This is sometimes called morphological data or clutter data in telecom tools.

What to require:

• Minimum 6 to 8 classes for rural/macro planning; 18 to 19 classes for urban 5G planning
• Resolution: 25 to 50 m (rural sub-1 GHz), 5 to 10 m (3.5 GHz), 2 to 5 m (mmWave)
• Classification delivered pre-coded for the target planning tool (Atoll, Planet, TEOCO)

What to avoid:

• Repurposing general-purpose land cover data (e.g., ESA WorldCover at 10 m with 11 classes) as telecom clutter without recoding: classification systems are not compatible
• Using 100 m resolution national land cover datasets for urban 5G planning

3. 3D Building Vectors: Mandatory for Urban 3.5 GHz and All mmWave

For any 5G deployment in an urban environment at 3.5 GHz or above, 3D building data transitions from optional to mandatory. This is the most important data upgrade from a 4G to a 5G geodata stack.

Why mandatory:

• LoS/NLoS classification cannot be performed deterministically without per-building geometry
• Rooftop diffraction calculations require exact building heights, not class-average heights
• FWA qualification (per-premise service eligibility) is impossible without per-building data
• Street canyon propagation modeling at 3.5 GHz requires building edge positions on both sides of the street

What to require:

• LOD1 minimum (footprint + height) for most 3.5 GHz and mmWave planning
• LOD2 (footprint + roof geometry) for dense urban mmWave where rooftop angle matters
• Building capture rate ≥90% (≥95% for mmWave FWA)
• Height RMSE ≤2 m (≤1 m for mmWave)
• Delivered in Shapefile or GeoPackage with height as a numeric attribute field

4. 3D Vegetation: Mandatory for Urban and FWA at 3.5 GHz+

Vegetation is one of the most significant and commonly underestimated propagation factors at 5G frequencies.

• At 3.6 GHz, including 3D tree data versus no tree data improved propagation prediction accuracy by 12.8% in a Barcelona validation study
• At 26 GHz, a single tree in the LoS path can cause 35.3 dB of path loss
• Averaged vegetation attenuation at 26 GHz is approximately 1 dB per meter of tree depth

For FWA at 3.5 GHz, a tree between the cell and the customer premises may be the deciding factor in whether that premises qualifies for service. 2D clutter classification cannot identify the specific tree. Only a vegetation layer with individual tree positions and heights can.

What to require:

• Individual tree polygons with height attributes (total height, trunk height, canopy height)
• Resolution sufficient to identify trees ≥3 m height
• Seasonal variation notation where leaf-on vs. leaf-off matters for the project timeline

The Recommended (Not Mandatory) Layers

Road Network Vectors

Road network data is not a direct propagation input but enables street-level antenna placement analysis for small cells, drive test route planning and visualization, navigation and addressing in field survey workflows, and accessibility analysis for candidate sites.

Standard format: Shapefile or GeoPackage with road classification attributes (highway, trunk, primary, secondary, residential).

Population Density Maps

Population density data is used for demand-driven capacity planning rather than propagation modeling. It answers the question: where should I place cells to maximize served users, not just geographic coverage?

Time-of-day population maps at 10 m resolution are particularly valuable for urban deployments where daytime commercial zone population differs significantly from nighttime residential population.

Digital Surface Model (DSM)

When 3D building vectors are not available, a high-resolution DSM serves as an approximation. A DSM at 0.5 to 1 m resolution captures building heights in aggregate. Planners use it for clutter height data and as a proxy for diffraction calculations. It is inferior to 3D building vectors for precise edge diffraction and LoS classification.

The Optional (Project-Specific) Layers

Indoor Building Layouts

For enterprise 5G private networks, dense urban indoor coverage planning, or iBwave-based indoor planning, floor plans or building interior geometry are needed. These are project-specific and have no globally sourced equivalent.

Wall and Fence Data

LuxCarta presented automated wall extraction from satellite imagery at IGARSS 2024, achieving 80.31% precision and 86.32% recall. Wall data is relevant for campus network planning, military applications, and any scenario where below-rooftop physical barriers affect small cell propagation. For most commercial telecom planning, wall data is optional.

Orthoimagery

High-resolution satellite or aerial imagery is not a propagation input but serves as a visual reference for site selection, field survey preparation, and stakeholder communications. Many operators already hold licensed imagery for their coverage areas.

Administrative Boundaries

Country, region, municipality, and postal district boundaries are used for reporting, license compliance, and coverage obligation documentation. They are not propagation inputs.

How to Prioritize Layers on a Budget

When data budget is constrained, the priority order for urban 5G planning is:

1. DTM: foundational, non-negotiable
2. LULC (18+ class, ≤10 m): without this, no propagation model can function
3. 3D building vectors (LOD1): mandatory for 3.5 GHz urban, highest accuracy impact
4. 3D vegetation (individual trees): critical for FWA and mmWave
5. Population density: for capacity planning and site prioritization
6. Road network: for field survey and site visualization
7. DSM: useful if 3D vectors are not yet available for all areas
8. LOD2 buildings: upgrade from LOD1 for mmWave dense urban
9. Orthoimagery: reference only

For rural sub-1 GHz planning, the priority list is much shorter: DTM, then 2D LULC, then roads.

How LuxCarta Addresses This

LuxCarta delivers the complete mandatory and recommended data stack for 5G network planning from a single satellite-based production pipeline. The company's core portfolio covers every mandatory layer: DTM (including HDTM), LOD1 and LOD2 3D building vectors (93%+ capture rate), LULC at 18 to 19 classes and 50 cm resolution, individual 3D vegetation with canopy and trunk separation, and time-of-day population density maps at 10 m resolution.

All layers are co-registered, covering the same geographic extent, with consistent coordinate reference systems and licensing. Delivered in standard GIS formats (SHP, GeoJSON, GeoPackage, GeoTIFF), all layers plug directly into Forsk Atoll, InfoVista Planet (including the AIM propagation model), TEOCO Asset, and iBwave for indoor planning without format conversion.

For operators who need to assess which layers provide the most value for their specific project, the BrightEarth platform allows trial extraction of individual layers for a pilot area, enabling direct before/after comparison in the planning tool before purchasing full-coverage data.

Frequently Asked Questions

Do I need all layers at the same resolution?

Not necessarily. The resolution of each layer should match the propagation model's sensitivity to that layer. A DTM at 5 m resolution and LULC at 5 m resolution are consistent. However, 3D building data is vector-based and inherently higher resolution than raster layers, so it is normal to have a 10 m DTM alongside 1 m-precision building vectors in the same project.

What happens if my building data and DTM have different coordinate systems?

This causes systematic spatial offsets: building footprints appear in the wrong location relative to terrain. All layers in a propagation dataset must use the same coordinate reference system (CRS), verified before import into the planning tool. Mixed CRS datasets are one of the most common sources of systematic prediction errors.

Can I use different vendors for different layers in the same project?

Yes, and this is common: a national DTM from a government source, commercial building data from a vendor, and LULC from another. The critical requirement is that all layers are co-registered: same CRS, consistent geographic extent, no systematic spatial offsets between layers. When ordering from multiple vendors, specify the CRS and tile schema explicitly to ensure consistency.

How do I know if my existing data stack is sufficient for a 5G upgrade?

Review each layer against the thresholds in this guide for your target frequency band and environment type. If you are upgrading from 4G to 5G at 3.5 GHz in an urban area, the most common gaps are: absence of 3D building vectors (you may have 2D clutter only), insufficient LULC class granularity, and no vegetation height data. These are the three layers most commonly added when transitioning from a 4G to a 5G planning dataset.

Are there any regulatory requirements for geodata in telecom licensing?

Regulatory requirements vary by country. Some national regulators require network coverage maps to be submitted in specified formats using standardized geodata. In the EU, the European Electronic Communications Code (EECC) influences how operators report coverage. Always verify with the national regulator whether specific geodata standards apply to your coverage reporting obligations.

LuxCarta provides AI-powered 3D geospatial data solutions for telecom, simulation, and smart city applications worldwide. Learn more at luxcarta.com or explore on-demand extraction at BrightEarth.