Coding and channel estimation for block fading channels

Wireless communication channels are commonly modeled by time-varying random processes that exhibit memory. A simple model is the block fading channel model. In this thesis, we derive a union bound on the performance of binary coded systems over block fading channels. Noncoherent and coherent receivers are considered with different assumptions on the channel side information (SI) at the receiver. We derive the union bound for Rayleigh, Rician and Nakagami distributed block fading channels. Systems employing single and multiple transmit antennas are considered. From the results, the tradeoff between channel diversity and channel estimation is investigated. Moreover, we study the effect of the parameters of the channel and the space diversity on the optimal channel memory.; As an effort to solve the channel estimation problem in multi-antenna transmission over block fading channels, we derive a pilot-aided iterative receiver for joint decoding and channel estimation. In the receiver, initial channel estimation is obtained using orthogonal pilot sequence insertion, and then soft information from the decoder is used to update the estimation. Results show that using 3 iterations in the iterative receiver results in a performance close to that of the best achievable performance.; Trellis space-time (ST) codes using the I-Q encoding scheme provide a large time diversity. For performance evaluation and decoding, the "super-trellis" of the composite code is necessary which is too complex in general. In this thesis, the performance of I-Q ST codes is analyzed using the transfer functions of the component codes. Moreover, two low-complexity iterative receivers for I-Q ST codes are proposed and compared to the optimal decoding in complexity and performance. Results show that using the iterative receivers with 3 iterations provide most of the coding gain of the optimal decoding.