Abstract:The efficient deployment and operation of any wireless communication ecosystem rely on knowledge of the received signal quality over the target coverage area. This knowledge is typically acquired through radio propagation solvers, which however suffer from intrinsic and well-known performance limitations. This article provides a primer on how integrating deep learning and conventional propagation modeling techniques can enhance multiple vital facets of wireless network operation, and yield benefits in terms of efficiency and reliability. By highlighting the pivotal role that the deep learning-based radio propagation models will assume in next-generation wireless networks, we aspire to propel further research in this direction and foster their adoption in additional applications.

What problem does this paper attempt to address?

The problem that this paper attempts to solve is the performance limitations faced by traditional radio propagation models when dealing with the complex and diverse propagation environments of next - generation wireless networks. Specifically, traditional propagation models have deficiencies in terms of accuracy, computational efficiency, and adaptability, and are unable to meet the requirements of new - generation wireless networks (such as millimeter - wave communication, reconfigurable intelligent surface (RIS), high - speed mobile communication, etc.). To solve these problems, the paper proposes to enhance multiple key aspects of wireless network operations by integrating deep learning (DL) and traditional propagation modeling techniques, thereby improving efficiency and reliability. ### Specific manifestations of the problem 1. **Complex and diverse propagation environments**: - New - generation networks support millimeter - wave and terahertz communications. These frequency bands are greatly affected by scattering, and near - field propagation mechanisms need to be considered. - Installing a reconfigurable intelligent surface (RIS) will change the propagation environment from a passive medium to an active entity, affecting communication quality. - In vehicle and high - speed railway communications, channel conditions change rapidly; satellite or unmanned aerial vehicle communications are affected by multiple environmental factors. 2. **Limitations of traditional models**: - **Empirical or statistical models**: Although they are easy to implement and have low complexity, they have low accuracy and are unable to capture complex propagation mechanisms. - **Deterministic models**: Based on Maxwell's equations, they have high accuracy but require a large amount of computational resources and are difficult to apply on a large scale. - **Full - scale adoption of deterministic methods** (such as ray tracing or finite - difference time - domain simulation) is limited by computational resources, resulting in a trade - off problem between accuracy and efficiency. ### Advantages of deep learning - **Enhanced accuracy**: Deep - learning models use complex nonlinear neural operators and are able to capture the complex dependencies of wireless channel behavior. - **Computational efficiency**: Thanks to GPU parallelization implementation, deep - learning models can quickly simulate wireless channel conditions. - **Data - driven**: There is no need to understand the specific mechanisms of radio wave propagation, and learning is directly from data. ### Main objectives of the paper By introducing a deep - learning - based radio propagation model, the paper aims to: - Improve the efficiency and accuracy of key applications such as wireless network planning, RIS simulation, user - equipment positioning, and uncertainty quantification. - Promote further research and facilitate the adoption of these models in more application scenarios. ### Application examples 1. **Network planning**: Use data - driven models to generate radio maps, significantly reducing optimization run - time. 2. **RIS modeling**: Map RIS configurations to their radiation patterns through deep - learning models to achieve real - time and efficient RIS configuration inference. 3. **Positioning**: Generate high - fidelity fingerprint databases to improve positioning accuracy and robustness. 4. **Uncertainty quantification**: Replace ray tracers, greatly reducing computational time. In summary, by combining deep - learning and traditional propagation - modeling techniques, the paper aims to solve the performance bottlenecks encountered by traditional models in complex and variable wireless environments and promote the development of the wireless communication ecosystem.

Empowering Wireless Network Applications with Deep Learning-based Radio Propagation Models

EM DeepRay: an Expedient, Generalizable, and Realistic Data-Driven Indoor Propagation Model

Cellular Network Radio Propagation Modeling with Deep Convolutional Neural Networks

A Machine Learning based Intelligent Propagation Model for RSRP prediction

Wireless Networks Design in the Era of Deep Learning: Model-Based, AI-Based, or Both?

Intelligent Propagation Model Method for RSRP Prediction Based on Machine Learning

Deep-Learning-Based Wireless Resource Allocation With Application to Vehicular Networks

Deep Learning based Wireless Resource Allocation with Application to Vehicular Networks

Deep Learning for Wireless Communications: An Emerging Interdisciplinary Paradigm

QoS and Jamming-Aware Wireless Networking Using Deep Reinforcement Learning

Application of End-to-End Deep Learning in Wireless Communications Systems

AI-assisted Agile Propagation Modeling for Real-time Digital Twin Wireless Networks

Deep Learning for Physical-Layer 5G Wireless Techniques: Opportunities, Challenges and Solutions

Deep Learning Frameworks for Cognitive Radio Networks: Review and Open Research Challenges

Deep Learning for Wireless Physical Layer: Opportunities and Challenges

Radio Map Prediction from Aerial Images and Application to Coverage Optimization

Deep Learning Challenges and Prospects in Wireless Sensor Network Deployment

A Data-Driven Approach for Designing Wireless Signal Propagation Model

PropEM-L: Radio Propagation Environment Modeling and Learning for Communication-Aware Multi-Robot Exploration

Machine Learning-Based Path Loss Modeling with Simplified Features

Development of Machine Learning-Based Radio Propagation Models and Benchmarking for Mobile Networks