Adaptive hybrid ARQ using convolutional and turbo codes on fading channels, and its application to the third-generation wireless communications

Fading in wireless environments is a major obstacle with AWGN (additive white gaussian noise) to reliable communication. In order to reduce the adverse effects of fading, channel coding is often used. In this dissertation, the performance of convolutional codes and turbo codes based on coherent BPSK signal over Rayleigh fading channels is examined. Three channel models are considered: memoryless, fast, and slow fading with AWGN.; The channel coding is combined with a new hybrid ARQ (automatic-repeat-request) technique to prevent losses of data frame from burst errors. It employs RCPC (rate compatible punctured convolutional) codes and RCPT (rate compatible punctured turbo) codes for frugal use of incremental redundancy. The new hybrid ARQ scheme adaptively adjusts the code rates of the first transmission according to the number of average retransmissions at given SNR (signal-noise-ratios) range. This dissertation shows that the adaptive hybrid ARQ yields better throughput than the type-2 hybrid ARQ for fast fading channels at long distances.; The performance of turbo codes used in third-generation Code Division Multiple Access (CDMA) 2000 is evaluated. This dissertation also shows performances of turbo codes by block and S-random interleaving techniques. Performance of adaptive hybrid ARQ technique is also studied in the CDMA 2000 environment.