A Tool for Power-Density Prediction of Radio Propagation in Urban Environments

The ability to predict transmission loss or field strength distribution is crucial for determining coverage in planning personal communication systems. Previous transmission loss prediction methods have considered specular reflection and diffraction of electromagnetic waves by and over buildings between the base station and the mobile terminals. These phenomena are of great importance in urban areas where there are many high buildings. But diffuse scattering is usually neglected. In urban environments, there are many irregularities such as windows, balconies, stucco, and so on in the outside building walls, and that are comparable with the wavelength of mobile communication (about 16.7cm for 1.8GHz). Hence we can suppose that the surface variations of most outside building walls are more than the threshold (2.79 cm for 1.8GHz and 45° incidence angle) at which roughness warrants consideration, real building walls can be thought as rough surfaces. Therefore it is reasonable to simplify the radio-wave propagation in urban environments to the electromagnetic scattering problem. According to this simplification, and utilizing advanced mathematical methods and three-dimensional (3D) graphics techniques, it was possible to develop a practical 3D prediction tool for urban areas, running on a PC. This thesis presents this promising and practical transmission loss prediction tool based on received radio power density. It allows to simulate very fast the average transmission loss (or mean field strength) distribution in urban radio propagation environments. The technique uses a 3D optical propagation model based on optical radiant flux transfer method (known to the computer graphics community as “radiosity”) and a simplified city-information database. The associated software product named “RadioPower” (version 1.10) has been developed with MSVC++ 5.0 under Windows 9X/NT/2000. It consists of pre-processing, field-strength simulation, 3D post-processing and visualization. There are about 17000 lines of source code. If the PC has 512MB memory and a powerful CPU (Pentium 1k), RadioPower allows to process the entire 3D map data of an European city, and to simulate and visualize the average transmission loss distribution over all the (wall and road) outside surfaces and the interesting points at one time with acceptable time-consumption and engineering-precision. The software offers much convenience in the actual design. As an sample, I have realized the simulation of average transmission loss distribution in the whole urban environment of Vienna ́s fourth district around Fakultät für Elektrotechnik und Informationstechnik. That utilizes a detailed city-map which contains location, shape and height of buildings. The simulated results are demonstrated and visualized in RadioPower (Typical run-time was about 10 minutes on a PC with the following specification: Windows 98/P-II 350MHz CPU/128M RAM). Comparison between the simulated results and the data actually measured previously demonstrate good coincidence (less than 5dB standard deviation of the error), which is an indication of the validity of RadioPower’s model and algorithms.