Generative AI on SpectrumNet: An Open Benchmark of Multiband 3D Radio Maps

TL;DR

This work introduces SpectrumNet, the largest open-source benchmark for multiband 3D radio maps that integrates terrain and climate information to address generalization gaps in radio map construction. It details the dataset design (3D spatial structure, five frequency bands, 11 terrain types, and 3 climate types) and provides baseline evaluations (UNet, CBAM, interpolation) across terrain, height, and frequency, highlighting where current methods fail and where generalization improves with richer data. The study demonstrates that 3D, multiband, and weather-aware data are essential for accurate radio-map construction and suggests concrete directions for dataset expansion, model enhancements, and cross-domain learning to advance SAGIN-enabled wireless networks. Overall, SpectrumNet serves as a comprehensive platform to train, evaluate, and generalize generative models for dense radio maps in diverse real-world conditions with potential impact on localization, planning, and network optimization.

Abstract

Radio map is an efficient demonstration for visually displaying the wireless signal coverage within a certain region. It has been considered to be increasingly helpful for the future sixth generation (6G) of wireless networks, as wireless nodes are becoming more crowded and complicated. However, the construction of high resolution radio map is very challenging due to the sparse sampling in practical systems. Generative artificial intelligence (AI), which is capable to create synthetic data to fill in gaps in real-world measurements, is an effective technique to construct high precision radio maps. Currently, generative models for radio map construction are trained with two-dimension (2D) single band radio maps in urban scenario, which has poor generalization in diverse terrain scenarios, spectrum bands, and heights. To tackle this problem, we provide a multiband three-dimension (3D) radio map dataset with consideration of terrain and climate information, named SpectrumNet. It is the largest radio map dataset in terms of dimensions and scale, which contains the radio map of 3 spacial dimensions, 5 frequency bands, 11 terrain scenarios, and 3 climate scenarios. We introduce the parameters and settings for the SpectrumNet dataset generation, and evaluate three baseline methods for radio map construction based on the SpectrumNet dataset. Experiments show the necessity of the SpectrumNet dataset for training models with strong generalization in spacial, frequency, and scenario domains. Future works on the SpectrumNet dataset are also discussed, including the dataset expansion and calibration, as well as the extended studies on generative models for radio map construction based on the SpectrumNet dataset.

Figures (15)

• Figure 1: Terrain scenarios in SpectrumNet dataset.
• Figure 2: Terrain and building information.
• Figure 3: Illustration for 3D map with terrain and building information.
• Figure 4: Building maps and radio maps on different heights.
• Figure 5: Radio maps of different bands in frequency domain.