RadioMap_Dataset_Reflection/Scattering_Counts
The dataset was generated using the open‑source TensorFlow‑based library Sionna for ray‑tracing simulations, covering reflection and scattering counts across different urban areas in China. The simulated area is a 620‑meter square, with both transmitters and receivers positioned at a height of 1.5 m. The dataset includes raw simulation output files, RSSI grayscale images, transmitter location images, and scene images.
Dataset description and usage context
RadioMap_Dataset_Reflection/Scattering_Counts
Dataset Overview
This dataset was produced by ray‑tracing (RT) simulations using the open‑source TensorFlow library Sionna. The simulation environment is based on various urban regions in China, focusing on a square area with a side length of 620 m. Transmitters (TXs) are uniformly distributed at a height of 1.5 m, and ground‑level receivers (RXs) are fixed at 1.5 m.
Simulation Parameters
- Frequency: 2.5 GHz
- Transmit Power: 23 dBm
- Transmit Height: 1.5 m
- Transmit Antenna Array: 8×4 elements, vertical and horizontal spacing 0.06 m (half‑wavelength), isotropic radiation pattern, dual‑polarized
- Receive Height: 1.5 m
- Receive Antenna Array: Single element, isotropic radiation pattern, cross‑polarized
Data Files
Scene_File
Contains 500 scene folders created and exported with Blender 3.6.
RadioMap_raw
Raw simulation output files generated by Sionna. Filenames range from 1_10.npy to 500_10.npy, indicating the scene number. Each file has a shape of 256×256×10×6, corresponding to (Rx_coordinates(i), Rx_coordinates(j), Tx_number, depth_max). Values represent RSSI. depth_max denotes the total number of reflections and scatterings; a depth_max of 0 indicates a scene without reflections or scatterings.
RSSI
Grayscale images derived from the RadioMap_raw files. Each point’s value is scaled to the range [0, 1] using the formula max{ (P_R - P_R,thr) / (P_R,max - P_R,thr) , 0 }, where P_R,max is the maximum received signal strength (‑13 dB) and P_R,thr is set to ‑70 dB. For example, 1_0_0.png represents the RadioMap for scene 1, TX 0, with depth_max = 0.
Tx
Grayscale images showing TX locations for each scene; pixel value 255 marks the TX position, all other values are 0. For instance, 1_0.png depicts the TX‑0 location in scene 1.
Scene
Grayscale images of the scenes obtained by a horizontal slice at 1.5 m through the 3D simulated environment. Pixel value 0 indicates building structures; all other values are 0.
Examples
Scene Images
| Scene 1 | TX 1_0 |
|---|---|
 |  |
RadioMap Images
| RSSI 1_0_0 | RSSI 1_0_1 | RSSI 1_0_2 |
|---|---|---|
 |  |  |
| RSSI 1_0_3 | RSSI 1_0_4 | RSSI 1_0_5 |
|---|---|---|
 |  |  |
In the examples, RSSI 1_0_0 denotes a scenario with no reflections or scatterings, RSSI 1_0_1 indicates a single reflection/scattering event, and so on.
Future Updates
Datasets with additional TX data for each scene are in preparation and will be released soon.
Pair the dataset with AI analysis and content workflows.
Once the source passes your review, move straight into summarization, transformation, report drafting, or presentation generation with the JuheAI toolchain.