A Study Of Radio Channel Prediction Based On Ray Tracing Algorithm For Outdoor And Indoor Scenarios

In the forthcoming years, the market for indoor and outdoor wireless networkingfacilities is expected to grow considerably in both commercial and domestic sectors.Thus, a thorough investigation of indoor and outdoor propagation channelcharacteristics represents a fundamental step toward the design and the implementationof such applications. This makes it necessary to have an accurate propagation model topredict propagation channel characteristics in indoor and outdoor environment.Traditionally, propagation models based on empirical and semi-empirical models havebeen used for large scenarios. However, as the traffic is constantly growing, the cellssize has decreased and current models cannot provide good predictions. As aconsequence, deterministic methods as ray tracing are a good alternative to be used. Raytracing is a commonly used computational method for site-specific prediction of theradio channel characteristics of wireless communication systems. The ray tracingtechnique inherently provides time delay and angle of arrival information for multipathreception conditions. So, the ray tracing can provide the reliable theory basis forradio-wave propagation prediction and network planning.According to the format storage of geometric information for both indoor picocellenvironment and street microcell environment, this dissertation presents a appropriatemodeling approach. On this basis, the successful prediction of radio channel is achievedthrough using the ray tracing technique in the two environment, then the acceleration ofcoverage prediction is systematically studied. The main contents and innovations are asfollows:1. Two theoretical foundations of ray tracing are introduced at first: geometricaloptics and uniform theory of diffraction. The main concepts and basic definitionsinvolved in the ray tracing algorithm in indoor picocell and street microcell aredescribed in detail. Then, the new virtual source tree is proposed. Moreover, all the adiochannel parameters predicted by ray tracing and the corresponding computationalformulas are also shown.2. A modeling approach in typical complex indoor environments is presented. And,a new storage format of environment information used in three-dimensional ray tracingmethod is designed. The ray-path classification technique for all the indoor propagationpaths is elaborated. Then, the operation procedure of coverage predition based on theproposed threed-dimensional ray tracing is represented in detail. The results indicatethat the proposed prediction model works well for both narrowband and wideband prediction for indoor picocell applications.3. Based on the study of the main idea of the three-demensional polar sweepalgorithm, some repeated judgments can be avoided. Through the analysis of theefficiency of virtual source tree creation, the method using power threshold control andlimit number is proposed to improve the efficiency of virtual source tree creation andavoid tracing the void ray paths. Considering the particularity of the propagationmechanism for diffraction and the relationship between receiver points and geometricinformation of surroundings, the corresponding data structure is designed, resulting inan improvement of the efficiency of intersection tests.4. A modeling approach for the geometric information in typical street microcellenvironments is presented. Also, the source of these environment information isintroduced. Utilizing the digital map to store the environment information of streetmicrocell, the main idea of the novel outdoor ray tracing model and the correspondingoperation procedure are described in detail. Furthermore, the implementation procedureof creating virtual source tree and the avoidance of unnecessary intersection tests aregiven. The results indicate that a good prediction accuracy for both narrowband andwideband prediction for indoor picocell applications is achieved.5. According to the characteristic of the quasi three-dimensional ray tracingalgorithm, intelligent preprocessing techniques for the geometric information ofbuildings are improved. During the process of creating the virtual source tree, someacceleration techniques in regard to reflection source are investigated. A very importantconcept orientation face set and improved polar sweep algorithm are used, solvingthe problem that the efficiency of creating the virtual source tree is not high in theoriginal ray tracing method. The handling of visible receivers for both reflection anddiffraction sources and the threshold control of each ray path are proposed. On this basis,the efficiency of coverage prediction over a large area is well improved.