Radiomap prediction improved with prior knowledge and residual learning
Rethinking Radiomap Blind Prediction with Limited Environment and Configuration Representations
Machine Learning
Summary
Predicting how radio signals spread in an area is difficult because important details about the environment and base stations are often missing. The authors show that the best possible prediction balances what can be learned from available data with uncertainty that can’t be removed. They introduce a new method that starts with a basic prediction guided by prior knowledge and then corrects it by learning the predictable errors. Their approach works better than traditional methods when predicting signal maps across different locations and base station setups.
What this means in practice
- •For wireless network planners: Generate accurate radio coverage maps in new environments without field measurement campaigns using combined prior models and learned corrections.
- •For urban delivery drone operators: Predict radio signal availability across various city layouts to improve drone communication reliability using environment-aware radiomap prediction.
Authors
Xiaojie Li, Yu Han, Han Fang, Shangqing Liu, Shi Jin, Chao-Kai Wen
Abstract
Radiomap blind prediction infers radiomaps from observable representations of the propagation environment and base station (BS) configuration without field measurements. These representations are inherently incomplete and cannot uniquely determine the target radiomap. Under squared loss, we identify the conditional-mean radiomap as the population-optimal deterministic target and decompose domain risk into target-approximation error and irreducible uncertainty. The train-test risk gap motivates propagation priors as cross-domain guidance, although their partial or simplified forms may bias the attainable predictor. We therefore propose RadioDecomp, which treats a prior-guided predictor as a correctable base and uses deterministic residual refinement to learn its remaining predictable discrepancy. We instantiate RadioDecomp as RadioLSR (LoS-Shadow-Residual). Experiments under cross-configuration and cross-environment settings show that RadioLSR is especially effective for cross-configuration generalization and provides overall gains over a controlled monolithic counterpart under cross-environment generalization.