An Exposimetric Electromagnetic Comparison of Mobile Phone Emissions: 5G versus 4G Signals Analyses by Means of Statistics and Convolutional Neural Networks Classification

Abstract

Overview

The study investigates the non-thermal biological effects of microwaves with a focus on mobile phone emissions. A comprehensive analysis of real-life exposure conditions under various parameters using peak exposure analysis is proposed as opposed to time-averaged analysis.

Methodology

• The study leverages advanced analytical tools such as signal and spectrum analyzers with adequate real-time analysis bandwidths to assess emissions from mobile phone applications.
• Statistical methods including the amplitude probability density (APD) function and the complementary cumulative distribution function (CCDF) were utilized along with channel power measurements and recorded spectrogram databases to differentiate emissions from 4G and 5G networks.

Findings

• Significant differences in electric field strengths were observed; 5G emissions exhibit on average a 60% higher electric field strength compared to 4G at a 10 cm distance from the phone. The highest disparities were during internet video streaming, with 5G displaying three times the field strength of 4G.
• Crest factors of 5G emissions were almost doubled during certain applications compared to 4G, indicating a more frequent prevalence of high power levels in 5G emissions.
• The processing capabilities of the YOLO v7 deep learning algorithm provided excellent recognition and classification rates for signals, enhancing our understanding of user exposures under different network conditions.

Conclusion

This work enhances the understanding of electromagnetic field exposure dynamics, suggesting noticeable variations between 4G and 5G technologies and providing insights into non-thermal effects of EMF, which remain a critical area for ongoing research.