analiza aiqbench
Point2Radio represents a significant methodological shift in wireless network simulation by leveraging foundation models to learn transferable propagation priors across diverse environments. This approach eliminates the need for computationally expensive per-scene meshing or explicit path tracing, enabling high-fidelity radio field prediction in milliseconds using only material-aware point clouds and transceiver queries. By achieving a 76.7% error reduction compared to standard baselines while supporting multiple output tasks through shared encoders, this work demonstrates how machine learning can fundamentally alter the efficiency of physical layer modeling for next-generation networks.
Point2Radio: A Foundation Model for Cross-Scene Radio Fields from Material-Aware Point Clouds
arXiv:2607.28994v1 Announce Type: cross Abstract: High-fidelity radio fields are typically simulated for every scene--transmitter configuration or fitted separately to each scene, failing to exploit propagation structures shared across environments. We present Point2Radio, a foundation model that learns a transferable propagation prior from multiple environments. Given a material-aware point cloud and a transmitter (TX) setting, a common encoder produces a TX-conditioned scene representation that can be queried at arbitrary receiver (RX) locations. Task-specific query decoders map this representation to different radio quantities, e.g., three-dimensional (3D) path-gain (PG) fields and power angular spectra (PAS). At inference for a new scene, the model uses only a material-aware point cloud and transceiver queries, running in milliseconds on a single GPU without meshes or explicit path tracing. We evaluate PG prediction on a scene-disjoint split of a 337-scene corpus containing 86,272 TX-conditioned fields. Point2Radio achieves 0.871 dB mean absolute error (MAE), reducing error by 76.7% relative to a same-split UNet-style baseline. The same encoder also supports PAS prediction via a task-specific decoder. Experiments further show that light target-scene fine-tuning improves adaptation to a specific environment.
Abstract:High-fidelity radio fields are typically simulated for every scene--transmitter configuration or fitted separately to each scene, failing to exploit propagation structures shared across environments. We present Point2Radio, a foundation model that learns a transferable propagation prior from multiple environments. Given a material-aware point cloud and a transmitter (TX) setting, a common encoder produces a TX-conditioned scene representation that can be queried at arbitrary receiver (RX) locations. Task-specific query decoders map this representation to different radio quantities, e.g., three-dimensional (3D) path-gain (PG) fields and power angular spectra (PAS). At inference for a new scene, the model uses only a material-aware point cloud and transceiver queries, running in milliseconds on a single GPU without meshes or explicit path tracing. We evaluate PG prediction on a scene-disjoint split of a 337-scene corpus containing 86,272 TX-conditioned fields. Point2Radio achieves 0.871 dB mean absolute error (MAE), reducing error by 76.7% relative to a same-split UNet-style baseline. The same encoder also supports PAS prediction via a task-specific decoder. Experiments further show that light target-scene fine-tuning improves adaptation to a specific environment.
Networking and Internet Architecture (cs.NI); Artificial Intelligence (cs.AI); Computer Vision and Pattern Recognition (cs.CV); Machine Learning (cs.LG)
arXiv:2607.28994 [cs.NI]
(or arXiv:2607.28994v1 [cs.NI] for this version)
https://doi.org/10.48550/arXiv.2607.28994 arXiv-issued DOI via DataCite (pending registration)
Text integral sub licența deschisă a publicației · 207 cuvinte
Subiecte conexe
Digital States 2026: AI Moves From Experiment to Infrastructure
govtech.com · 24 sept. 2026
2026.5: We're on the same frequency now 📡
Home Assistant · 6 mai 2026
2026.7: Automations that speak your language
Home Assistant · 1 iul. 2026
Radiance-Field Guided Pretraining: Scaling Localization Models with Unlabeled Wireless Signals
arXiv · 24 sept. 2026