Abstract

Overview

With the rapid expansion of 5G networks, concerns around electromagnetic field exposure are increasingly critical. However, existing methods to measure and reconstruct these fields in specific regions are limited.

Findings

• This study introduces the use of a conditional generative adversarial network (GAN) as a novel solution to accurately reconstruct electromagnetic field exposure maps in urban outdoor environments.
• The model leverages only a few sensors and depends heavily on the topological characteristics of the environment.
• It was compared against traditional methods like simple kriging, demonstrating superior accuracy and effectiveness in mapping.

Conclusion

The conditional GAN framework established in this research presents a promising approach to enhance electromagnetic field mapping and could be pivotal for public health considerations as 5G deployment increases.