In urban and dense suburban environments where future cells will be reduced in size, accurate propagation prediction in microcells will play an important role. Such propagation prediction for the line-of-sight (LOS) path in microcells is already well developed. However, the situation for the non-LOS (NLOS) path, such as that around corners and parallel streets, is much less well understood....[ Read more ]

In urban and dense suburban environments where future cells will be reduced in size, accurate propagation prediction in microcells will play an important role. Such propagation prediction for the line-of-sight (LOS) path in microcells is already well developed. However, the situation for the non-LOS (NLOS) path, such as that around corners and parallel streets, is much less well understood.

In this thesis an investigation into techniques for improving NLOS microcell propagation prediction is provided. Specifically, two approaches are investigated, one of which is based on ray tracing whilst the other is based on empirical modelling. In the ray tracing approach, the formalism of a high frequency electromagnetic approximation known as the Uniform Theory of Diffraction is applied to the microcell environment. Rays reflected from the building walls and road in conjunction with diffraction from corners are utilised to give an estimate of the field strength. In the empirical approach, the already well established LOS model is adapted to incorporate additional corner loss factors and also side street path loss for NLOS propagation prediction.

The computer implementation of these models is also detailed and simulation results are also provided. These are used to illustrate the effects of corners, side streets and material parameters on the propagation characteristics.

Finally, comparisons of the simulation results to actual DCS 1800 microcell propagation measurement results from Hong Kong are made. These allow the accuracy of our models to be established.