| Simple and accurate channel models are desirable for effective analytical or simulation-based performance evaluation that is crucial for designing, developing, and optimizing wireless communication systems and network protocols. This dissertation, for the first time, analytically develops the following simple theoretical channel models that accurately capture useful channel statistics in modern wireless networks: (1) A multipath channel model for the time variations of both line-of-sight and non-line-of-sight fixed indoor communication (FIC) channels in short-range or uncrowded indoor wireless networks; (2) Two Block Markov models (BMMs) for the error processes and the vacation (error or pause) processes of the packet transmission of adaptive modulation systems (AMSs) in very slow and flat fading channels; (3) A BMM for the packet transmission error processes of multicarrier modulation systems (MMSs) in very slow and frequency-selective fading channels with independent fading for all subcarrier transmissions.;The BMMs are simple, tractable, and accurate channel models for performance analysis of network protocols working over the AMSs or the MMSs. The BMMs are shown to accurately predict the packet transmission statistics of the AMSs and the MMSs such as, the packet error rate and the average number of packets per error burst. Besides their simple and tractable Markov structures, the BMMs have both stationary state probabilities and the transition probabilities defined based on only packet transmission statistics. Therefore, the BMMs conveniently enable analytical and fast packet-level simulation-based performance evaluation that avoids directly dealing with physical-layer and propagation information.;Based on the multipath channel model, this dissertation derives a correct expression for FIC channel's Doppler spectrum useful for waveform-level FIC channel simulations. Additionally, this dissertation derives the following FIC channel's statistics useful for FIC system design parameters: the level crossing rate, the average fade duration, the expected number of crossings of random phase per second, the probability density function of random frequency modulation, and the expected number of crossings of random frequency modulation per second. Finally, we analytically validate FIC channel's key ergodicity properties used for performance evaluation: the mean- and covariance- ergodicities of the inphase and the quadrature components of received signals, and the mean-ergodicity of received signal powers. |
