Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

PINN and GNN-based RF Map Construction for Wireless Communication Systems

Lizhou Liu, Xiaohui Chen, Zihan Tang, Mengyao Ma, Wenyi Zhang

Abstract

Radio frequency (RF) map is a promising technique for capturing the characteristics of multipath signal propagation, offering critical support for channel modeling, coverage analysis, and beamforming in wireless communication networks. This paper proposes a novel RF map construction method based on a combination of physics-informed neural network (PINN) and graph neural network (GNN). The PINN incorporates physical constraints derived from electromagnetic propagation laws to guide the learning process, while the GNN models spatial correlations among receiver locations. By parameterizing multipath signals into received power, delay, and angle of arrival (AoA), and integrating both physical priors and spatial dependencies, the proposed method achieves accurate prediction of multipath parameters. Experimental results demonstrate that the method enables high-precision RF map construction under sparse sampling conditions and delivers robust performance in both indoor and complex outdoor environments, outperforming baseline methods in terms of generalization and accuracy.

PINN and GNN-based RF Map Construction for Wireless Communication Systems

Abstract

Radio frequency (RF) map is a promising technique for capturing the characteristics of multipath signal propagation, offering critical support for channel modeling, coverage analysis, and beamforming in wireless communication networks. This paper proposes a novel RF map construction method based on a combination of physics-informed neural network (PINN) and graph neural network (GNN). The PINN incorporates physical constraints derived from electromagnetic propagation laws to guide the learning process, while the GNN models spatial correlations among receiver locations. By parameterizing multipath signals into received power, delay, and angle of arrival (AoA), and integrating both physical priors and spatial dependencies, the proposed method achieves accurate prediction of multipath parameters. Experimental results demonstrate that the method enables high-precision RF map construction under sparse sampling conditions and delivers robust performance in both indoor and complex outdoor environments, outperforming baseline methods in terms of generalization and accuracy.

Paper Structure

This paper contains 10 sections, 14 equations, 5 figures, 2 tables, 1 algorithm.

Table of Contents

  1. Introduction
  2. Multipath System Model
  3. RF Map Construction via PINN and GNN
  4. PINN Architecture
  5. GNN Architecture
  6. Joint Construction of RF Map via PINN and GNN
  7. Simulation and Performance Evaluation
  8. Simulation Scenario and Setup
  9. Simulation Results
  10. Conclusion

Figures (5)

  • Figure 1: PINN-GNN architecture for RF map construction.
  • Figure 2: Simulation scenarios: (a) DeepMIMO dataset, (b) USTC campus environment, (c) 3D model of USTC campus buildings.
  • Figure 3: Comparison of true and predicted distributions of the first-path parameters in S2: The first row illustrates the true power, delay, elevation angle, and azimuth angle, respectively, while the second row shows the corresponding predicted results.
  • Figure 4: CDF of the errors for the four parameters under different methods in S2: (a) power, (b) delay, (c) elevation angle, and (d) azimuth angle.
  • Figure 5: Channel impulse responses under different methods: (a) at location (-18.59, 18.67, 1) in S1, and (b) at location (160.52, 187.27, 1) in S2.