File size: 3,755 Bytes
9ac7a76 b11d8a4 a7c5a52 b11d8a4 a7c5a52 b11d8a4 a7c5a52 b11d8a4 a7c5a52 | 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 | ---
license: mit
---
# Dataset Card — Site-Specific MIMO Channel Generation via Diffusion and Flow Matching: Fidelity, Efficiency, and Downstream Utility
#### Link to paper (to be updated): [[TBC](TBC)]
#### Authors: Sina Beyraghi, Masoud Sadeghian, Firdous Bin Ismail, Angel Lozano, Paul Almasan, and Giovanni Geraci
Contact: Sina Beyraghi (<mohammadsina.beyraghi@telefonica.com>), Paul Almasan
## Abstract
This paper explores the use of generative models to synthesize high-quality, site-specific multiple-input multiple-output (MIMO) channel data, addressing the high cost of the extensive measurement campaigns required to acquire real-world data for AI-native wireless networks. Two location-conditioned generative paradigms are compared: a conditional denoising diffusion implicit model (cDDIM), and a conditional flow matching model (cFMM). Both these models generate MIMO channel matrices conditioned on user coordinates, to preserve the spatial structure of the deployment site. The approaches are evaluated across three dimensions: statistical fidelity (including beam consistency and effective rank), generation efficiency, and utility in downstream tasks such as channel-state information compression and beam alignment. Results across diverse propagation scenarios (28 GHz and 3.5 GHz, both line-of-sight and non-line-of-sight) demonstrate that both models accurately capture site-specific characteristics, even when trained on scarce ground-truth data. Notably, cFMM achieves a quality comparable to cDDIM with roughly an order of magnitude less inference time. Augmenting scarce site-specific datasets with these synthetic channels yields hefty performance gains in downstream physical layer tasks compared to using scarce data alone or stochastic channels.
---
## Files
| File | Frequency | Scenario | Size |
|---|---|---|---|
| `Final_Single_Scene_Channel_Sionna_V1_3_5GHz_LoS.npz` | 3.5 GHz | LoS only | — |
| `Final_Single_Scene_Channel_Sionna_V1_3_5GHz_NLoS.npz` | 3.5 GHz | NLoS only | — |
| `Final_Single_Scene_Channel_Sionna_V1_28GHz_LoS.npz` | 28 GHz | LoS only | — |
---
## Data format
Each `.npz` file contains a single array under the key `combined_array`:
```
shape: (N, N_rx, 1, N_tx, 1, 1, 4)
dtype: complex64
last dimension:
[0] — complex channel coefficient H
[1] — UE x-coordinate (metres)
[2] — UE y-coordinate (metres)
[3] — UE z-coordinate (metres)
```
To extract the channel matrix and UE coordinates from a file:
```python
import numpy as np
npz = np.load("Final_Single_Scene_Channel_Sionna_V1_3_5GHz_LoS.npz")
data = npz["combined_array"][:, :, 0, :, 0, 0, :] # (N, N_rx, N_tx, 4)
H = data[:, :, :, 0] # complex channel matrices, shape (N, N_rx, N_tx)
coords = data[:, 0, 0, 1:] # UE (x, y, z) positions, shape (N, 3)
```
---
## Generation
The datasets were generated with [NVIDIA Sionna RT](https://nvlabs.github.io/sionna/), a GPU-accelerated ray tracing engine for wireless channel simulation, over a single outdoor urban scene. Generation scripts and instructions are available in the [code repository](https://github.com/Telefonica-Scientific-Research/GenAI_Channel_Modeling/tree/main/Channel_Sionna_RT_Github).
---
## Downloading
```bash
git clone https://huggingface.co/datasets/PaulAlm/GenAI_Channel_Modeling_Datasets
```
Due to file size this may take several minutes. Individual files can also be downloaded manually from the Hugging Face web interface.
---
## Related resources
- **Code repository:** [GenAI_Channel_Modeling](https://github.com/Telefonica-Scientific-Research/GenAI_Channel_Modeling)
- **Pre-trained models:** [GenAI_Channel_Modeling_Models](https://huggingface.co/PaulAlm/GenAI_Channel_Modeling_Models)
---
## Citation
TBC.
|