Machine Learning-Based Path Loss Modeling with Simplified Features
Jonathan Ethier, Mathieu Chateauvert
TL;DR
This work tackles the challenge of accurate path loss prediction in non-LOS settings by using simple environmental features derived from path profiles. It evaluates three ML architectures—Log-Reg, XGBoost, and FCN—across three feature configurations that incorporate distance $d$, frequency $f$, and obstacle-related depths $o$, $t$, and $c$, with $o = t + c$. Training uses a rigorous city-holdout scheme on the OFCOM drive-test data, and results show that adding obstacle depth substantially lowers RMSE (by ~5 dB) and that FCN with 3–4 features achieves RMSE in the $6$–$8$ dB range, approaching physics-informed baselines while remaining computationally efficient. The study demonstrates that obstacle-depth modeling captures key propagation effects, yielding practical, environment-aware path loss predictions suitable for wireless network planning, and it identifies limitations and avenues for future improvements such as clutter classification and seasonal effects.
Abstract
Propagation modeling is a crucial tool for successful wireless deployments and spectrum planning with the demand for high modeling accuracy continuing to grow. Recognizing that detailed knowledge of the physical environment (terrain and clutter) is essential, we propose a novel approach that uses environmental information for predictions. Instead of relying on complex, detail-intensive models, we explore the use of simplified scalar features involving the total obstruction depth along the direct path from transmitter to receiver. Obstacle depth offers a streamlined, yet surprisingly accurate, method for predicting wireless signal propagation, providing a practical solution for efficient and effective wireless network planning.