Network Planning For Urban Microcells With Ray Tracing And Genetic Algorithms

In order to solve the contradiction between frequency spectrum resource andsystem capacity in mobile communication systems, cell radius becomes smaller andsmaller, and thus microcells and picocells have become popular. Typical microcellsusually lie in urban building canyon or indoor scenarios, the height of transmitter andreceiver antennas is usually lower than the surrounding buildings, where thetraditional statistical channel models are not proper. Therefore, ray tracing model forradio propagation characteristics prediction of microcells has become a hot topic,while network planning methods based on ray tracing have attracted great attention.This thesis studies a scheme which combines ray tracing and genetic algorithm tosolve network planning problem in urban microcells. Ray tracing can be used topredict radio propagation characteristics of microcells accurately. Afterwards, geneticalgorithm can be used to find the specific locations of base stations in urbanmicrocells. Owing to the shortcomings existing in traditional genetic algorithm, suchas massive calculation, dull later search and premature convergence problems, thisthesis adopts an improved genetic algorithm in literature. The adopted algorithmimproves traditional methods in the aspects of encoding method choice, fitnessfunction design, genetic operator improvement. Simulation results based on urbanmicrocells show that, the scheme which combines ray tracing and genetic algorithmcan effectively solve the base station site planning problem in urban microcells.Furthermore, in order to explicitly demonstrate abundant and abstract radiopropagation characteristics of urban scenarios, visual software for wireless networkplannning in urban environments is developed. In the macrocell scenarios, based ongeographic information data, the software utilizes stastical model to realize coverageanalysis of single base station, single frequency network and multiple frequencynetwork; In the microcell scenarios, combined with open graphic library (OpenGL),the software uses3D ray tracing model to achieve the visualization of radiopropagation characteristic parameters, such as field strength, time delay, angle, et al.Also, the adopted genetic algorithm is integrated in the wireless network planning software to solve base station site planning problem. Simulation results show that thewireless network planning software facilitates the acquisition of propagation statisticcharacteristics, displays the results of base station site planning in the form ofvisualization, and provides an intuitive and effective reference for the wirelessnetwork planning and optimization.