Radiomap prediction improves under incomplete propagation information
Radiomap Blind Prediction under Incomplete Observation: Error Characterization and Correctable Propagation-Prior Learning
Artificial IntelligenceMachine Learning
Summary
Predicting radio signal strength over an area usually needs measurements, but sometimes we must guess from incomplete information. The authors explain how missing data causes certain unavoidable errors and how to best aim predictions at average values. They propose a new method called RadioDecomp that starts with a physics-based guess and then learns to correct its mistakes. Tests show this method improves predictions when the initial assumptions are imperfect.
What this means in practice
- •For wireless network planners: Estimate radio coverage maps in new environments without costly field measurements by improving predictions from incomplete data about surroundings and base stations.
- •For telecommunications engineers: Enhance radio signal modeling tools by integrating corrections to physics-based predictions, reducing errors when only partial environmental data is available.
Authors
Xiaojie Li, Yu Han, Han Fang, Shangqing Liu, Guangxu Zhu, Shi Jin, Chao-Kai Wen
Abstract
Radiomap blind prediction aims to infer radiomaps from observable representations of the propagation environment and base station configuration without field measurements. In practice, the observable representations are inherently incomplete. Thus, the target radiomap is not fully determined by the inputs when generalizing to unseen configurations or environments. Under incomplete observation, we establish a population-level theory of deterministic radiomap blind prediction that identifies the conditional mean as its optimal target and separates prediction error into reducible predictor approximation and irreducible uncertainty caused by missing physical information. The framework further characterizes the train-test risk gap and the uncertainty reduction enabled by observation enrichment. Building on it, we reveal the dual role of propagation priors: they provide physically grounded guidance, yet their implementable forms may bias the attainable predictor. This motivates RadioDecomp, which treats a prior-guided predictor as a correctable base and learns its remaining predictable discrepancy through residual refinement. To evaluate RadioDecomp across distinct propagation-prior designs, we instantiate it with a feature-guided monolithic base and a LoS-Shadow structured base, yielding RadioFR and RadioLSR, respectively. Across random, cross-configuration, and cross-environment settings, experiments confirm the benefit of propagation-related representations and show that both instantiations improve upon their respective bases. Further controlled studies on base capacity, training-support coverage, and observation coarsening corroborate the proposed analysis.