Abstract:Radio maps quantify magnitudes such as the received signal strength at every location of a geographical region. Although the estimation of radio maps has attracted widespread interest, the vast majority of works rely on simulated data and, therefore, cannot establish the effectiveness and relative performance of existing algorithms in practice. To fill this gap, this paper presents the first comprehensive and rigorous study of radio map estimation (RME) in the real world. The main features of the RME problem are analyzed and the capabilities of existing estimators are compared using large measurement datasets collected in this work. By studying four performance metrics, recent theoretical findings are empirically corroborated and a large number of conclusions are drawn. Remarkably, the estimation error is seen to be reasonably small even with few measurements, which establishes the viability of RME in practice. Besides, from extensive comparisons, it is concluded that estimators based on deep neural networks necessitate large volumes of training data to exhibit a significant advantage over more traditional methods. Combining both types of schemes is seen to result in a novel estimator that features the best performance in most situations. The acquired datasets are made publicly available to enable further studies.

What problem does this paper attempt to address?

### What problems does this paper attempt to solve? This paper aims to solve the problems of effectiveness and performance evaluation of Radio Map Estimation (RME) in practical applications. Specifically, the paper mainly focuses on the following aspects: 1. **Limitations of existing research**: - Most studies on RME rely on simulated data and cannot verify the effects and relative performance of existing algorithms in real - world environments. - The understanding of the RME problem itself is not in - depth enough, especially in terms of performance and comparison on real - data. 2. **Filling the gaps in practical applications**: - The paper conducts a comprehensive and rigorous empirical study by collecting a large amount of actual measurement data to evaluate the performance of different RME algorithms in real - world environments. - The research uses two different technical means to collect data: one is a custom - made system based on Software - Defined Radio (SDR), and the other is data from 4G LTE networks. 3. **Evaluating and comparing existing algorithms**: - Four performance indicators are used to extensively analyze and compare existing RME algorithms, including traditional methods and methods based on Deep Neural Networks (DNN). - Empirical research shows that even with a relatively small number of measurement points, the estimation error is also relatively small, which proves the feasibility of RME in practical applications. 4. **Proposing a new hybrid estimator**: - Combining the advantages of traditional methods and DNN, a new type of hybrid estimator is proposed and shows better performance in most cases. 5. **Public data set**: - The collected data set is made public for other researchers to further study and verify. ### Summary The main objective of the paper is to conduct a comprehensive empirical study on the radio map estimation problem by using actual measurement data, evaluate the effectiveness of existing algorithms, and propose improvement plans. This not only fills the gaps in existing research but also provides valuable data resources and references for future research.

Radio Map Estimation: Empirical Validation and Analysis

Theoretical Analysis of the Radio Map Estimation Problem

Radio Map Estimation: A Data-Driven Approach to Spectrum Cartography

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

Radio Map Estimation Using a Generative Adversarial Network and Related Business Aspects

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

Spectrum Surveying: Active Radio Map Estimation with Autonomous UAVs

Quantized Radio Map Estimation Using Tensor and Deep Generative Models

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

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

Radio Map Estimation with Deep Dual Path Autoencoders and Skip Connection Learning

Radio Map Prediction from Aerial Images and Application to Coverage Optimization

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

RADIANCE: Radio-Frequency Adversarial Deep-learning Inference for Automated Network Coverage Estimation

Constructing Indoor Region-based Radio Map without Location Labels

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

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

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

Geo2SigMap: High-Fidelity RF Signal Mapping Using Geographic Databases

RadioUNet: Fast Radio Map Estimation with Convolutional Neural Networks

Engineering Radio Map for Wireless Resource Management