Research On Wireless Propagation Fading Indoor Model Under Various Conditions

In recent years-mobile communications are in a period of rapid development.With the emergence of various new services and more mobile users who are increasingly demanding high-quality communications,network optimization has become an inevitable key research topic.Since the indoor environment is quite complicated,it is very important to obtain a model that can accurately estimate and predict the attenuation of indoor signal propagation.A propagation model like this can not only predict the signal propagation and get the signal coverage of a specific area,but also provide a theoretical basis for future network optimization.This article first introduces various indoor models,including the commonly used empirical models and theoretical models.The empirical models include free space models which is the most basic,two-ray models,and the models of Keenan-Motley.The theoretical models are mainly include Ray tracking mode and time domain finite difference model.The main calculation formulas and scope of application of each model are outlined.This includes the applicable frequency bands and corresponding scenarios,And this article analyzes the advantages and disadvantages of various models.Then,various fitting methods are described,including traditional linear fitting and nonlinear fitting,such as classical least squares method,linear interpolation,Gauss-Newton nethod and so on.An artificial neural network method used in the case of large data volumes is also described.And the advantages and disadvantages of several fitting methods are compared and analyzed.Taking a set of real data as an example,various fitting methods are used for actual fitting.And then the fitting method is optimized.That is to say,the two methods:smoothing spline and linear fitting are combined.The smoothing spline method is used to average the data,and the interference values which are excessively high and low are rernoved,and the data after smoothing spline uesd as original data which is linearly simulated.It is to obtain the exact function relation expression while ensuring the validity of the data.After the verification process,it is found that the model has large errors in the real environment.For the problem that the error exist between the path loss and the true path loss in the actual application of indoor model predict,an improved model which adapt to the situation of passing walls and multi-person and glass door scenes are proposed.The model is based on the empirical model proposed by CYLee.The planar model of the actual indoor structure is established.The vector antenna analyzer and portable spectrum analyzer are used to transmit a certain power and different frequency of frequency signal,by changing the receiving antenna positiol to obtain test result That is,a practical test of three kind of scenarios is performed using the CW test method.The C.YLee theoretical model was validated and optimized using smooth splines and 1.near fitting methods.The obtained simulation results show that the improved model reduces the error in the actual measured loss,and the modified model can better predict the path loss of the corresponding scene,the root mean square error(RMSE)between the measured value and the predicted value of the improved model is less than 8dB,which is in accordance with the model calibration results analysis criteria.Finally,the distribution of the signal path loss of the corrected model in the real scene is analyzed in a gridded manner,and the three-dimensional picture of the test site is drawn.The effect of the improved model in practical application is further verified.The improved model provides a realistic reference for signal propagation loss prediction in a specific scenario and has reliability,accuracy and analyzabilily data.The practicality of the model was confirmed.