No description available.

Thesis (M.S.)--Ohio University, August, 2002. / Title from PDF t.p.

No description available.

Thesis (Ph. D.)--University of Washington, 2000. / Vita. Includes bibliographical references (leaves 169-184).

No description available.

No description available.

No description available.

No description available.

No description available.

The increased application of digital terrestrial microwave radio links in communication networks has renewed attention in techniques of estimating the probability of multipath fading distributions. Nevertheless, the unpredictable variation of the wireless transmission medium remains a challenge. It has been ascertained that the refraction of electromagnetic waves is due to the inhomogeneous spatial distribution of the refractive index, and causes adverse effects such as multipath and diffraction fading. The knowledge of the characteristics of such causes of these fading phenomena is essential for the accurate design of terrestrial line of sight (LOS) links of high performance and availability. Refractivity variation is random in space and time and cannot be described in a deterministic manner and has to be considered as a random variable with probabilistic characteristics. In this dissertation, radiosonde soundings data is used in characterizing the atmospheric conditions and determining the geoclimatic factor K used in predicting the distribution of multipath fading for five locations in South Africa. The limitations of radiosonde measurements are lack of time resolution and poor spatial resolution. The latter has been reduced by spatial interpolation techniques in our study, specifcally, the Inverse Distance Weighting (IDW) method. This is used in determining the point refractivity gradient not exceeded for 1 % of the time from which the geoclimatic factor is estimated. Fade depth and outage probability due to multipath propagation is then predicted from the International Telecommunications Union Recommendations (ITU-R) techniques. The results are compared with values from Central Africa. The results obtained using the ITU-R method are also compared with region-based models of Bannett-Vigants of USA and Morita of Japan. Three spatial interpolation techniques (Kriging, Thin-Plate Spline and Inverse Distance Weighting) are then used in interpolating the geoclimatic factor K in places where radiosonde data is not available. The estimated values have been used to develop contour maps for geoclimatic factor K for South Africa. Statistical assessment of these methods is done by calculating the root mean square error (RMSE) and the mean absolute error (MAE) between a set of control points and the interpolated results. The best performing method is used to map the seasonal geoclimatic factor K for the entire study region. The estimated values of geoclimatic factor will improve accuracy in predicting outage probability due to multipath propagation in LOS links in the region which is a key contribution of this work. / Thesis (M.Sc.Eng.)-University of KwaZulu-Natal, Durban, 2013.