Modeling and wideband characterization of radio wave propagation in microcells
MetadataShow full item record
This dissertation focuses on radio wave propagation prediction modeling in microcellular environments. Mathematical modeling techniques and channel characterization have been addressed. The research begins with a survey of ray-tracing based propagation prediction modeling methods and the evolution of diffraction modeling, and makes three new contributions. First, diffraction theory has been used to enhance the accuracy in shadowed regions for ray-tracing algorithms by a novel, computationally fast approach. A parametric formulation of the UTD diffraction coefficient has been proposed using the inverse problem theory. Significant improvement in the estimation of the diffracted field for right-angle dielectric wedges is achieved over existing heuristic methods. This construction offers a faster estimation of the diffracted field and is suitable for real-time wireless channel estimation. The diffraction modeling method has been applied in indoor and outdoor environments to expedite propagation prediction. Next, the ray-tracing prediction modeling method developed on the NSF Montage project is applied and validated with published measurements at 800 MHz in the San Francisco financial area to investigate the ability to predict and model wave propagation in a dense urban environment. Cross-correlation of the predicted and the measured signal strengths has been calculated at various street intersections and correlation coefficients of higher than 0.8 are achieved under all circumstances where corner effect occurs. Further investigations show that the waveguide effect in the street canyon is only applicable to the propagation phenomena in radial streets with respect to the transmitter. While in the case of cross streets that are not close to the transmitter, wave penetration through the building walls makes significant contribution to signal variations. Finally, site-specific prediction is applied to the campus of Pickle Research Center at UT-Austin to study the MIMO channel characteristics. The prediction using a 4 × 8 uniform linear array system operating at 1.8 GHz shows that S 4W is able to reproduce fading statistics and that the estimation of MIMO channel capacity using ray-tracing method is dependent on the accuracy of the input propagation environment.