Abstract

Overview

The rapid growth of mobile networks has raised concerns about the health effects of electromagnetic (EM) radiation from base stations. Traditionally, evaluations of EM radiation levels have relied on assessments of maximum power output which often overlook traffic load considerations.

Methodology

This study introduces an artificial intelligence (AI) and machine learning (ML) driven approach to optimize GSM base station EM radiation estimates using actual traffic data. Researchers collected EM radiation measurements and traffic data from various GSM base stations selected based on their location and configuration.

Techniques Used

• Linear regression
• Random forests
• Neural networks
• Unsupervised learning for clustering base stations by radiation profile

Findings

The application of AI and ML models demonstrated improved prediction accuracy for EM radiation levels when compared to traditional power-based methods. The study highlighted the high generalizability of the AI/ML model across different types of base stations. It also revealed that site-specific factors are influential in daily EM radiation patterns post-clustering.

Implications

The findings support the use of traffic data in monitoring EM radiation which can aid telecom operators and regulatory bodies in analyzing EM radiation more accurately and effectively. This adjustment leads to optimized public safety practices and enhanced trust in mobile network operations.

Future Directions

The researchers suggest the inclusion of 5G networks and small cell network enhancements in future studies, along with the integration of these methodologies into smart city platforms.

Conclusion

This AI/ML methodology enhances the understanding and management of EM radiation, proposing substantial improvements in public safety and data-driven strategies for future telecom infrastructures.