| In recent years,people's need for a better quality of communication is growing rapidly.How to apply the existing communication technology to better serve people without increasing the communication cost has been the goal of communication operators,at the same time the issue trouble communication researchers deeply.Therefore,the optimization of network has become an inevitable focus Research topics.5G network has become a new generation of communication network.Compared with the previous 2G,3G and 4G networks,5G network uses higher frequency and wider bandwidth.5G network not only increases the system capacity,but also improves the communication efficiency,which means the accuracy of the communication attenuation prediction model is required to be higher.The better the prediction accuracy of the model,the more the planning of the wireless network.The significance of this thesis lies in how to improve the accuracy of model prediction.In this thesis,Firstly,the three mechanisms of radio signal propagation,reflection,diffraction and scattering,as well as the related theoretical calculation,are briefly described,which provides the basis for introducing the theoretical prediction model of microcellular.Secondly,the main reasons of signal attenuation in the process of channel propagation are analyzed,including multipath delay,obstacle block and propagation distance,Doppler frequency shift,etc,as well as attenuation selective shadow fading flat fading,frequency selective fading and space selective fading.Finally,it introduces the classic microcellular prediction model,describes the corresponding classic microcellular prediction model to adapt to the scene and the corresponding scene parameters,and analyzes the advantages and disadvantages of each model.In this thesis,the method to improve the prediction accuracy of the model is to continuously add scene factors to make the prediction model adapt to the needs of the scene,so as to achieve the purpose of improving the prediction accuracy of the model.The main thing in this thesis is to analyze and study the three typical scenes in the microcellular prediction model,which are:buildings,terrain,direction of the street,the three scenes to signal transmission the loss is corrected.in the building loss model,the area proportion and distance are taken as the scene parameters,and the Lee microcellular building loss model constructed by the least square method is compared with the building loss model constructed by the RF and GBRT algorithms;in the terrain scene,the effective antenna height is taken as the parameter,and the altitude difference between the two receiving and transmitting points is calculated by three methods respectively,and the comparative analysis is obtained in the street trend scene,the incident angle and distance are used as the scene correction parameters,and the RF and GBRT algorithm are used to fit the data for comparative analysis,and the best performance algorithm is selected to predict the loss of signal propagation prediction.The validation results of data fitting of three scenarios show that:the prediction loss results of the modified scenario model are highly consistent with the actual test values,and the maximum error,RMSE and R~2 between the measured value and the predicted value all conform to the model discrimination criteria.Finally,the scene prediction model is integrated into the microcellular model.It is verified that 80%of the error of the integrated improved Lee microcellular model is within 5dB The maximum error is less than 7 dB.Compared with the original Lee microcellular model,the prediction accuracy of the model is improved by about6 dB,and the adaptability of the model to the scene is greatly enhanced.By the end of the thesis,aiming at the scene parameters extracted in the experiment and the different algorithms used in data fitting,the gain and loss in the experiment and data processing are analyzed,the correction results of different scenes are summarized,and the next work plan and prospect are put forward. |
PDF Full Text Request