Total: 1
Conventional base station (BS) deployments typically prioritize coverage, quality of service (QoS), or cost reduction, often overlooking electromagnetic field (EMF) exposure. Whereas EMF exposure triggers significant public concern due to its potential health implications, making it crucial to address when deploying BS in densely populated areas. To this end, this paper addresses minimizing average EMF exposure while maintaining coverage in a 3D urban scenario by jointly optimizing BS deployment and power. To address this, firstly, accurate EMF prediction is essential, as traditional empirical models lack the required accuracy, necessitating a deterministic channel model. A novel least-time shoot-and-bounce ray (SBR) ray-launching (RL) algorithm is therefore developed to overcome several limitations of current simulators and is validated with real-world measurements. Secondly, to further reduce computational complexity, unlike using a fixed grid size to discretize the target area, the adaptive grid refinement (AGR) algorithm is designed with a flexible grid to predict the overall EMF exposure. Finally, based on the EMF exposure predictions, the Nelder-Mead (NM) method is used in the joint optimization, and urban user equipment (UE) distributions are incorporated to better reflect real-world conditions. When evaluating the benefits of the whole process, the results are compared against using empirical channel models, revealing notable differences and underestimation of EMF exposure that highlight the importance of considering real-world scenario.