Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

RadioGen3D: 3D Radio Map Generation via Adversarial Learning on Large-Scale Synthetic Data

Junshen Chen, Angzi Xu, Zezhong Zhang, Shiyao Zhang, Junting Chen, Shuguang Cui

TL;DR

This work proposes an efficient data synthesis method to generate high-quality 3D radio map data and proposes a 3D model training scheme based on a conditional generative adversarial network (cGAN), yielding a 3D U-Net capable of accurate RME under diverse input feature combinations.

Abstract

Radio maps are essential for efficient radio resource management in future 6G and low-altitude networks. While deep learning (DL) techniques have emerged as an efficient alternative to conventional ray-tracing for radio map estimation (RME), most existing DL approaches are confined to 2D near-ground scenarios. They often fail to capture essential 3D signal propagation characteristics and antenna polarization effects, primarily due to the scarcity of 3D data and training challenges. To address these limitations, we present the RadioGen3D framework. First, we propose an efficient data synthesis method to generate high-quality 3D radio map data. By establishing a parametric target model that captures 2D ray-tracing and 3D channel fading characteristics, we derive realistic coefficient combinations from minimal real measurements, enabling the construction of a large-scale synthetic dataset, Radio3DMix. Utilizing this dataset, we propose a 3D model training scheme based on a conditional generative adversarial network (cGAN), yielding a 3D U-Net capable of accurate RME under diverse input feature combinations. Experimental results demonstrate that RadioGen3D surpasses all baselines in both estimation accuracy and speed. Furthermore, fine-tuning experiments verify its strong generalization capability via successful knowledge transfer.

RadioGen3D: 3D Radio Map Generation via Adversarial Learning on Large-Scale Synthetic Data

TL;DR

This work proposes an efficient data synthesis method to generate high-quality 3D radio map data and proposes a 3D model training scheme based on a conditional generative adversarial network (cGAN), yielding a 3D U-Net capable of accurate RME under diverse input feature combinations.

Abstract

Radio maps are essential for efficient radio resource management in future 6G and low-altitude networks. While deep learning (DL) techniques have emerged as an efficient alternative to conventional ray-tracing for radio map estimation (RME), most existing DL approaches are confined to 2D near-ground scenarios. They often fail to capture essential 3D signal propagation characteristics and antenna polarization effects, primarily due to the scarcity of 3D data and training challenges. To address these limitations, we present the RadioGen3D framework. First, we propose an efficient data synthesis method to generate high-quality 3D radio map data. By establishing a parametric target model that captures 2D ray-tracing and 3D channel fading characteristics, we derive realistic coefficient combinations from minimal real measurements, enabling the construction of a large-scale synthetic dataset, Radio3DMix. Utilizing this dataset, we propose a 3D model training scheme based on a conditional generative adversarial network (cGAN), yielding a 3D U-Net capable of accurate RME under diverse input feature combinations. Experimental results demonstrate that RadioGen3D surpasses all baselines in both estimation accuracy and speed. Furthermore, fine-tuning experiments verify its strong generalization capability via successful knowledge transfer.
Paper Structure (24 sections, 29 equations, 9 figures, 4 tables)

This paper contains 24 sections, 29 equations, 9 figures, 4 tables.

Table of Contents

  1. Introduction
  2. System Model
  3. Communication Model
  4. Problem Formulation
  5. Efficient Data Synthesis and cGAN-based Radio Map Generation
  6. Efficient Synthesis of Radio3DMix Dataset
  7. Feature Vector Construction
  8. Label Construction via Data Synthesis
  9. Properties of Radio3DMix Dataset
  10. cGAN-based 3D Radio Map Estimation
  11. Model Selection
  12. Data Pre-processing
  13. cGAN-based Model Training
  14. Experimental Results
  15. Experimental Configurations
  16. ...and 9 more sections

Figures (9)

  • Figure 1: Diagram of the proposed RadioGen3D framework.
  • Figure 2: Illustration of 3D channel fading under vertical polarization.
  • Figure 3: Visualized horizontal slices at different heights. The transmitter is located at height of $z=10$ m.
  • Figure 4: Visualized vertical slices with different model coefficients.
  • Figure 5: The proposed cGAN-based training process in RadioGen3D
  • ...and 4 more figures

Theorems & Definitions (1)

  • Remark 1: Effects of 3D Channel Fading and Antenna Polarization