An Overview of Machine Learning Techniques for Radiowave Propagation Modeling

Abstract

We give an overview of recent developments in the modeling of radiowave propagation, based on machine learning algorithms. We identify the input and output specification and the architecture of the model as the main challenges associated with machine learning-driven propagation models. Relevant papers are discussed and categorized based on their approach to each of these challenges. Emphasis is given on presenting the prospects and open problems in this promising and rapidly evolving area.

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

“…Until now, ML has been employed to tackle a wide range of wireless network design-related problems, such as resource allocation, user mobility analysis, localization, and wireless channel modelling [7]- [9]. The latter case has recently attracted significant interest, as radio propagation modelling is the cornerstone of the cellular network design [10], [11].…”

“…Computational efficiency usually appears as a bottleneck for ray tracing, due to the substantial simulation time and memory required to trace all the ray paths, when the number of scattering objects and ray intersections within the simulated space increases. Data-driven approaches aim to alleviate this limitation by integrating ray tracing simulators with ML algorithms which are capable of learning and inferring radio propagation parameters [10], [11]. In particular, artificial neural networks (ANNs) have been widely used in an effort to expedite [19] or even replace ray tracing simulators [20].…”

On Machine Learning Techniques Toward Path Loss Modeling in 5g and Beyond Wireless Systems

“…More recently, with the motivation of reducing computational demands while maintaining good accuracy, machine learning methods have emerged as promising tools for radio propagation prediction in the different 5G application scenarios and communication environments [40].…”

“…However, many ML algorithms have been regarded as black-box systems because of the lack of knowledge of the machine's internal mechanism to output the prediction after training. Therefore, despite the several studies for PL modeling based on ML techniques, it is still unclear why the ML models improve over the traditional PL models, how to interpret them, and how well the trained model can generalize to unseen data [32,40].…”

“…In addition, the conduction of generalization tests is necessary to provide a reliable conclusion about an ML model's generalization capacity. The term generalization capacity denotes the ability of the model to accurately predict the output value when the instance is unknown during the training of the model [40,70].…”