| The topics discussed in this thesis can be placed into three categories: channel modeling, channel estimation, and performance prediction.; Regarding channel modeling, first we propose the spherically invariant random process model for multipath fading in terrestrial channels. Then we discuss the application of the gamma distribution for shadow fading in terrestrial and satellite channels. In addition, we propose a novel model for the composite multipath/shadow fading in satellite channels. Afterwards, we derive a parametric correlation model for single-input single-output mobile fading channels. We then extend that model to a general spatio-temporal correlation function for multiple-input multiple-output mobile fading channels in multiantenna systems. All the proposed models are validated either by real data or published experimental results in the literature. Moreover, the utility of the models are demonstrated through numerous system analysis/design examples.; After channel modeling, we move on to channel estimation. Specifically, first we derive simple moment-based estimators for the parameters of Rice and Nakagami fading distributions. Then, we propose an estimator for the speed of a mobile user in cellular systems, based on the rate of maxima of the signal envelope. Finally, we devise a method for joint estimation of the angle of arrival and angle spread of a stationary user in cellular systems, using an antenna array. In addition to the theoretical analysis, the performance of the estimators are evaluated using Monte Carlo simulations.; At the end, we study the dynamic characteristics of diversity systems in multipath fading channels and derive a new formula for the associated level crossing rate. |
