Research On Electromagnetic Radiation Prediction Method Of 5G Base Station

In recent years,people have paid more and more attention to the electromagnetic radiation safety of 5G base stations with the continuous development of 5G communication technology.Due to the increase in the 5G frequency band,in order to ensure the quality of communication,the density and transmitting power of 5G base stations have increased。The 5G base station adopts Massive MIMO technology,which can achieve 3D beam-forming,making the electromagnetic radiation prediction of previous generations of base stations being significant overestimation of the 5G base station.Therefore,in order to reduce people’s worries,it is necessary to theoretically predict the electromagnetic radiation level of 5G base stations,and provide help for the actual measurement of electromagnetic radiation levels of 5G base stations.This paper predicts the electromagnetic radiation of 5G base stations based on the ray-tracing algorithm.This paper presents a ray-tracing algorithm suitable for urban environments.Diffraction ray tracing is introduced into this method,which solves the problems of complex diffraction tracing and large amount of calculation in urban environment.This paper analyzes the influencing factors of electromagnetic radiation of 5G base stations,mainly including mobile phone working mode and user distribution,and obtains base station antenna radiation for various scenarios through Matlab simulation.When using the ray-tracing model to predict,firstly import the buildings and other environments around the base station,and then import the antenna patterns of each scene separately,and finally obtain the electromagnetic radiation prediction results of the 5G base station.And compared with the measured results to verify the accuracy of this method.This paper predicts the electromagnetic radiation of 5G base stations based on the random forest model.In this paper,the measurement principle of 5G base stations is analyzed,and the theoretical maximum radiation point is calculated according to the antenna height,beam width and antenna down-tilt angle.Several 5g base stations in Beijing are actually measured,and the measurement results are taken as the data set.In this paper,the electric field strength of the maximum radiation point、the horizontal and vertical distance from the maximum radiation point、the measurement time and the data download duration are input into the random forest model as features,and the electric field strength of the predicted point is output.The method in this paper is compared with the KNN model and BP neural network,and the results verify the performance superiority of the method in this paper.