Content area
Abstract
In this dissertation, concatenated coding techniques combined with iterative decoding techniques are studied and applied to partial response channels.
In Chapter 2, a soft-output decoding algorithm, the a posteriori probability algorithm, is reviewed.
In Chapter 3, turbo coding and decoding for channels without intersymbol interference is reviewed. Concepts essential to turbo decoding such as concatenated coding and pseudo-random interleaving will be introduced.
Turbo decoding for an ideal partial response channel, where the channel is the precoded partial response polynomial followed by additive white Gaussian noise, is discussed in Chapter 4. Several competing architectures are discussed and compared. Performance of these architectures as a function of precoder, outer code, partial response target, interleaver, block-size, and number of iterations is addressed.
Turbo decoding for more realistic magnetic recording channel models is discussed in Chapter 5. The simple white noise channel model is replaced with an equalized Lorentzian channel model to ascertain the performance with the introduction of colored noise. Media noise is introduced with the microtrack channel model. The bit error performance and the byte error burst statistics are investigated for systems employing the various channel models.
In Chapter 6, a novel decision-feedback equalization (DFE) architecture, termed parallel DFE, is introduced. Performance of this alternative to conventional PRML systems is investigated. A turbo decoding technique for use with a parallel DFE detector is presented.
