Uncertainty Estimation for Path Loss and Radio Metric Models
This work provides scalable uncertainty estimation for wireless network modeling, offering potential insights for network planning and operations, though it is incremental as it applies an existing method to new data.
The research tackled uncertainty estimation in machine learning-based radio metric and path loss models using Conformal Predictive Systems, achieving statistically robust 95% confidence intervals that generalized across cities like Toronto, Vancouver, and Montreal with high coverage and reliability, and reduced RMSE as dataset difficulty decreased.
This research leverages Conformal Prediction (CP) in the form of Conformal Predictive Systems (CPS) to accurately estimate uncertainty in a suite of machine learning (ML)-based radio metric models [1] as well as in a 2-D map-based ML path loss model [2]. Utilizing diverse difficulty estimators, we construct 95% confidence prediction intervals (PIs) that are statistically robust. Our experiments demonstrate that CPS models, trained on Toronto datasets, generalize effectively to other cities such as Vancouver and Montreal, maintaining high coverage and reliability. Furthermore, the employed difficulty estimators identify challenging samples, leading to measurable reductions in RMSE as dataset difficulty decreases. These findings highlight the effectiveness of scalable and reliable uncertainty estimation through CPS in wireless network modeling, offering important potential insights for network planning, operations, and spectrum management.