Abstract
The problem of the efficient use of Hybrid Automatic-Repeat-reQuest (Hybrid-ARQ)
in wireless communication has attracted a considerable amount of research. In this
thesis, the use and implementation of turbo codes as the Forward Error Correction
(FEC) code for Hybrid-ARQ is investigated. The major accomplishments of the
research include both the analysis and implementation of turbo Hybrid-ARQ. The
thesis begins by obtaining a tractable bound for the performance of turbo codes with
M-ary Quadrature-Amplitude-Modulation (M-ary QAM). The research considers the
design problem of turbo coded Hybrid-ARQ optimized for AWGN and fading channels.
The design problem of turbo Hybrid-ARQ in wideband channels is considered
and an optimization strategy is proposed based on Orthogonal-Frequency-Division-
Multiplexing (OFDM). The research also presents a novel rate scalable encoder structure
that optimal selects a disparate but optimal pair of component codes given the
channel conditions. A second part of the thesis considers the implementation of turbo
Hybrid-ARQ in Very Large Scale Integration (VLSI ) systems. A design for a single
architecture for Type-I and Type-II turbo Hybrid-ARQ is suggested in addition to
approaches for improving performance of the Soft-Output-Viterbi-Algorithm(SOVA)
decoder core. The research also proposes a SOVA decoder architecture that exploits
reliability information to select between the SOVA and bi-directional SOVA.