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.

What problem does this paper attempt to address?

The problem that this paper attempts to solve is the challenges faced in constructing high - precision radio maps. Specifically, the paper points out two main problems: 1. **Sparse Sampling Rate**: In actual systems, the sampling rate of radio sensors is very low, usually less than 1%, which means that more than 99% of the radio map information cannot be directly obtained through radio sensing technology. 2. **Limited Effective Bandwidth**: The effective bandwidth of a radio map depends on the working bandwidth of radio sensors. In order to cover the signal information of common wireless systems from very low frequency (VLF) to Ka - band, the cost of deploying radio sensors is extremely high. To solve these problems, researchers propose using generative artificial intelligence (Generative AI) to construct high - precision radio maps. The generative model can utilize sparsely - sampled signals and the knowledge of the generative model to create synthetic data to fill in the gaps in real - world measurements, thereby obtaining high - resolution broadband radio maps. In addition, the existing radio map datasets have the following limitations: - **Single Terrain Scene**: Existing datasets mainly focus on urban scenes and ignore other terrain scenes (such as rural and mountainous areas), which weakens the generalization ability of the generative model in different terrain scenes. - **Lack of Terrain Information**: Existing datasets only consider the influence of objects on the ground (such as buildings and trees) on radio propagation, while ignoring factors such as geographical height and terrain materials that have a significant impact on radio propagation. - **Lack of Three - Dimensional Spatial Information**: Existing datasets are mainly two - dimensional and do not contain information at different heights, which limits the research on non - ground networks (such as air - space - ground integrated networks). - **Limited Frequency Range**: Existing datasets mainly focus on the C - band, which limits the performance of the generative model in the frequency domain. To solve the above problems, the author proposes a multi - band three - dimensional radio map dataset named SpectrumNet. SpectrumNet is currently the largest radio map dataset, covering 11 different geographical scenes, 3 climate scenes, 5 different frequency bands, and providing three - dimensional spatial information at three different height levels. Through this dataset, the generative model can be better trained to improve its generalization ability in the spatial, frequency, and scene domains. ### Summary The main goal of this paper is to solve the problems of sparse sampling and limited bandwidth encountered in constructing high - precision radio maps by introducing the SpectrumNet dataset, and to improve the generalization ability of the generative model in different terrains, frequencies, and application scenarios.

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

Supreme: Fine-grained Radio Map Reconstruction Via Spatial-Temporal Fusion Network

Model and Transfer Spatial-Temporal Knowledge for Fine-Grained Radio Map Reconstruction

Three-Dimensional Radio Spectrum Map Prediction Based on Fully Connected Neural Network

ACT-GAN: Radio map construction based on generative adversarial networks with ACT blocks

Two-Stage Radio Map Construction with Real Environments and Sparse Measurements

High-Efficiency Urban 3D Radio Map Estimation Based on Sparse Measurements

ACT‐GAN: Radio map construction based on generative adversarial networks with ACT blocks

RadioGAT: A Joint Model-based and Data-driven Framework for Multi-band Radiomap Reconstruction via Graph Attention Networks

Deep Spectrum Cartography: Completing Radio Map Tensors Using Learned Neural Models

Spectrum Surveying: Active Radio Map Estimation with Autonomous UAVs

Distributed Radio Map Modeling Based on Feature Augmentation of Augmented Scatter Graph

Quantized Radio Map Estimation Using Tensor and Deep Generative Models

Exemplar-Based Radio Map Reconstruction of Missing Areas Using Propagation Priority

A FL-Based Radio Map Reconstruction Approach for UAV-Aided Wireless Networks

Model-Free Radio Map Estimation in Massive MIMO Systems via Semi-Parametric Gaussian Regression

Radio Map Estimation -- An Open Dataset with Directive Transmitter Antennas and Initial Experiments

Fast and Accurate Cooperative Radio Map Estimation Enabled by GAN

RME-GAN: A Learning Framework for Radio Map Estimation based on Conditional Generative Adversarial Network

Generative AI for RF Sensing in IoT systems

3D Spectrum Mapping and Reconstruction under Multi-Radiation Source Scenarios