Super-Resolution Experimental Validation and Polarimetric Extension of the Effective Roughness Diffuse Scattering Models
Fuente:
arXiv
Saved in:
| Main Authors: | , , , , , , |
|---|---|
| Format: | Preprint |
| Published: |
2026
|
| Subjects: | |
| Online Access: | |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1866913172844183552 |
|---|---|
| author | Melloni, Giacomo Chuang, Jack Berweger, Samuel Vitucci, Enrico M. Degli-Esposti, Vittorio Gentile, Camillo Golmie, Nada |
| author_facet | Melloni, Giacomo Chuang, Jack Berweger, Samuel Vitucci, Enrico M. Degli-Esposti, Vittorio Gentile, Camillo Golmie, Nada |
| contents | The experimental validation of diffuse scattering models has long been limited by the inability to spatially separate specular and diffuse contributions in measured channels. This paper overcomes this limitation by combining super-resolution multipath component (MPC) extraction, which resolves individual propagation paths including the specular component, with digital-twin-assisted geometry, enabling the spatial separation of specular and diffuse contributions from bistatic measurements at 28~GHz. Using this framework, we provide the first measurement-driven validation of the Effective Roughness (ER) model with independent characterization of diffuse scattering across ten common building materials, each measured over 266 angular configurations and all polarization combinations (HH, HV, VH, VV). Furthermore, we extend the ER framework by proposing a novel angle-dependent cross-polarization discrimination (XPD) model, capturing the geometry-dependent nature of depolarization that is neglected in existing approaches. The proposed method reproduces the measured diffuse power trends, achieving RMSE values as low as 3 dB across the tested materials, and improves XPD prediction over the baseline constant-XPD model for nearly all material-polarization cases. These results establish a physically consistent and practically viable approach for high-fidelity channel modeling in mmWave systems. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2605_31267 |
| institution | arXiv |
| publishDate | 2026 |
| record_format | arxiv |
| spellingShingle | Super-Resolution Experimental Validation and Polarimetric Extension of the Effective Roughness Diffuse Scattering Models Melloni, Giacomo Chuang, Jack Berweger, Samuel Vitucci, Enrico M. Degli-Esposti, Vittorio Gentile, Camillo Golmie, Nada Signal Processing The experimental validation of diffuse scattering models has long been limited by the inability to spatially separate specular and diffuse contributions in measured channels. This paper overcomes this limitation by combining super-resolution multipath component (MPC) extraction, which resolves individual propagation paths including the specular component, with digital-twin-assisted geometry, enabling the spatial separation of specular and diffuse contributions from bistatic measurements at 28~GHz. Using this framework, we provide the first measurement-driven validation of the Effective Roughness (ER) model with independent characterization of diffuse scattering across ten common building materials, each measured over 266 angular configurations and all polarization combinations (HH, HV, VH, VV). Furthermore, we extend the ER framework by proposing a novel angle-dependent cross-polarization discrimination (XPD) model, capturing the geometry-dependent nature of depolarization that is neglected in existing approaches. The proposed method reproduces the measured diffuse power trends, achieving RMSE values as low as 3 dB across the tested materials, and improves XPD prediction over the baseline constant-XPD model for nearly all material-polarization cases. These results establish a physically consistent and practically viable approach for high-fidelity channel modeling in mmWave systems. |
| title | Super-Resolution Experimental Validation and Polarimetric Extension of the Effective Roughness Diffuse Scattering Models |
| topic | Signal Processing |
| url | https://arxiv.org/abs/2605.31267 |