OpenWorld Radio Twin

Under review · Code released under Apache 2.0
OpenWorld Radio Twin explorer over central Boston with a Sionna RT received-power map draped onto the 3D scene
Central Boston in the interactive explorer: three transmitters, 2,581 Overture buildings, and a Sionna RT received-power map over a 750 m radius draped on the scene. Downloads keep the exact solver cells and raw linear arrays.

Demonstration

0:00 / 7:01

Only the import changes

A geographic scene loads like a Sionna file scene. The returned object is a native sionna.rt.Scene; arrays, radio devices, solvers, radio maps, rendering and preview stay the upstream API.

# Sionna file scene
from sionna.rt import load_scene
scene = load_scene(filename)

# Geographic scene: same call, WGS84 coordinates
from openworld_radio_twin.rt import load_scene
scene = load_scene(latitude=52.3762, longitude=4.8993, radius_m=250)

From a point on the globe to a Sionna scene

Environmental layers from open geodata, compiled into one metric frame with the receiver surface and the radio outputs.

Building parts, terrain and semantic surfaces compiled into one native Sionna scene
1

Compile

Buildings from Overture and OpenStreetMap worldwide, from 3DBAG and Berlin LoD2 where they cover the site, or from Boston BPDA models by ID; bare-earth terrain from USGS 3DEP or Mapzen; water levelled per component; ground, vegetation and paved classes with ITU-R P.2040 materials, and fresh and sea water with the ITU-R P.527 water model.

2

Simulate

The scene is Mitsuba XML and PLY. Sionna RT solves radio maps on a terrain-following measurement surface and channel responses with PathSolver. The explorer places transmitters inside the compiled scene, by hand or from antenna sources, and drapes the result on its terrain; a CPU preview writes the same outputs.

3

Generate

One scene expands into seeded radio maps, each with its own transmitter layout and solver settings. Datasets resume after interruption, extend deterministically, and record the source, release and resolution behind every array.

What each layer changes

Explorer views and controlled comparisons; each comparison changes one input.

Exported products from one Boston run

Path gain export
Path gain (dB)
Received signal strength export
Received power (dBm)
SINR export
SINR (dB)
Transmitter association export
Transmitter association
Receiver-height voxel occupancy slice
Receiver-height voxel slice
Maximum occupied voxel elevation preview
Voxel z-depth preview

One 1,500 m × 1,500 m scene, three 3.5 GHz sector transmitters, 1 m cells, 1.5 m receiver plane. PNGs repeat each solver cell over a whole number of pixels, with transparent no-data; raw arrays ship alongside.

Features

Continuous globe

A CesiumJS world with display terrain everywhere. Drag a diagonal to define a scene; the compiled scene replaces the globe inside its square, with its own terrain, surfaces and buildings, and transmitters are placed inside it.

Buildings at source detail

Overture and OSM extrusions worldwide; native LoD2 roofs from 3DBAG and Berlin, selected automatically by location, and Boston's BPDA models by explicit ID.

Auditable terrain

Bare-earth DEMs from USGS 3DEP and Mapzen Skadi with source, model and local elevation arrays kept apart; water levelled and recorded.

Semantic surfaces

Non-overlapping ground, vegetation, paved and water meshes with documented ITU material proxies, each switchable at solve time.

Sionna-native

Terrain-following measurement surfaces, native propagation controls, path solvers and rendering on the same scene. The CPU preview writes the same outputs.

Reproducible datasets

Seeded expansion with PCG64, per-radio-map provenance, resume after interruption, deterministic extension, 16-bit packed path gain with optional PNG previews, release packaging with splits and checksums, a Parquet index and a PyTorch dataset reader.

Regional data and contributions

By default, a scene takes the openly licensed regional source that covers its centre, and worldwide sources elsewhere; worldwide antennas need an OpenCelliD token.

Buildings

  • Netherlands 3DBAG LoD2.2, CC BY 4.0
  • Berlin Berlin LoD2, Datenlizenz Deutschland – Zero – Version 2.0
  • Boston BPDA models, selected by identifier: no licence stated
  • Elsewhere Overture Maps, OpenStreetMap as fallback, ODbL

Antennas

  • France ANFR register, Licence Ouverte 2.0
  • Netherlands Antennebureau list, CC0 1.0
  • Poland UKE permits, CC BY 4.0
  • Australia ACMA register, selected by identifier: its licence restricts licensee details
  • Elsewhere OpenCelliD with the user's token, CC BY-SA 4.0

Terrain and surfaces

  • United States USGS 3DEP, public domain
  • Elsewhere Mapzen Terrain Tiles, terms of each upstream source
  • Surfaces Overture Maps worldwide, ODbL

Rules every regional source follows

  • Open terms for automatic routing. By default, a source is chosen by location only when its publisher lets anyone use, adapt and share the data for any purpose, with at most attribution and share-alike conditions. Other sources, such as the Boston models and the ACMA register, are selected by identifier.
  • Official boundaries. Routing tests the scene centre against official administrative boundaries published under open licences, which ship with the code together with their source and licence.
  • No silent fallback. An explicitly selected source serves the scene or reports why it cannot; another source never takes its place.
  • Recorded provenance. Every scene and dataset records the source and its version, the data release where the source names one, its licence and the credit to give.
  • Data stay with their providers. Apart from the routing boundaries, a few values in test fixtures and what screenshots, renders and notebook outputs show, the repository holds no provider data: data are downloaded at run time and cached locally, and a token is used for its download only and never recorded.

Contributing guide Source plugins

Get started

# install
git clone https://github.com/zhiheng-yang/openworld-radio-twin.git
cd openworld-radio-twin
./scripts/setup.sh
conda activate owrt

# interactive explorer
./scripts/run.sh          # http://127.0.0.1:8765

# compile a reusable scene
owrt scene compile --lat 52.3762 --lon 4.8993 --radius 250 --output scenes/amsterdam
from openworld_radio_twin import rt as owrt
from sionna.rt import PlanarArray, RadioMapSolver, Transmitter

scene = owrt.load_scene(latitude=52.3762, longitude=4.8993, radius_m=250)
scene.tx_array = PlanarArray(num_rows=1, num_cols=1, pattern="iso", polarization="V")
scene.rx_array = PlanarArray(num_rows=1, num_cols=1, pattern="iso", polarization="V")
scene.add(Transmitter(name="tx", position=owrt.position(
    scene, latitude=52.3770, longitude=4.9003, height_agl=25.0)))

radio_map = RadioMapSolver()(
    scene, measurement_surface=owrt.measurement_surface(scene, cell_size=(5.0, 5.0)))

Read the quickstart

BibTeX

@software{yang2026openworldradiotwin,
  author  = {Zhiheng Yang},
  title   = {OpenWorld Radio Twin: Interactive World-Scale Wireless Digital Twins},
  year    = {2026},
  version = {0.3.0},
  url     = {https://github.com/zhiheng-yang/openworld-radio-twin},
  license = {Apache-2.0}
}