The Research On Path Loss Prediction Of Radio Wave Propagation In Urban Areas

The study of radio wave propagation makes a significant impact on the planning, design andoptimization of wireless network and has been one of the hot research topics. Among the study ofradio wave propagation, the prediction of radio wave propagation path loss plays an important role.With the rapid increase of mobile communication users and the demand of the faster communicationrate, micro-cell and Pico cell system is widely adopted and the area of cell coverage is smaller andsmaller, which leads to the disappear of statistics similarity between cells and the invalidation of theconventional empirical prediction model based on measured data. Therefore, in this paper, we make adetailed study on the deterministic model like ray-tracing model and Gaussian beam tracing model,which is not only of theoretic value but also of broad prospect.First of all, statistical model is studied and the shortcoming of the model is that it must bemodified according to similar scenes. The deterministic model like ray tracing is studied in detail, andproposed a reverse ray tracing method based on visible walls. What’s more, numerical simulation isapplied in Ottawa city scene, of which the result shows that the prediction error is less than10dB.Secondly, as to the shortcoming of ray tracing like initial ray division, reception ball andcomplex diffraction, a new predicting model based on Gaussian beam tracing method is adopted. Theexpansion of source field, the representation of Gaussian beam with complex ray, the expansionparameters of complex ray, the propagation of Gaussian beams and the computation of the receivedfield are specifically studied. Simulation in the same scene as the made with ray-tracing method isalso carried out and the result is compared with that of ray-tracing method, which shows that theGaussian beam tracing model is simpler and as accurate as ray tracing method.Finally, as to the shortcoming of Gaussian beam-tracing method like time consuming, someimproved method like space dividing and key region selection is proposed. Simulation in somedifferent scenes is also made to verify the improved Gaussian beam tracing model. The results showthat the Gaussian beam tracing method is general in different scenes.