Publications-Detail

Turbo DeCodulation: Iterative Joint Source-Channel Decoding and Demodulation

Authors:
Clevorn, T.
Editors:
Vary, P.
Ph. D. Dissertation
 
School:
IND, RWTH Aachen
Adress:
Templergraben 55, 52056 Aachen
Series:
Aachener Beiträge zu Digitalen Nachrichtensystemen (ABDN)
Number:
24
Date:
Nov. 2006
ISBN:
3-861-30650-6
Language:
English

Abstract

The fundamental framework for today's digital communication systems is given by the information theoretical considerations of Shannon. It was soon discovered that due to the constraints of real world applications a joint design of the dirent components of the communication system, e.g., source coding, channel coding, and modulation, is benecial if not essential. A breakthrough in closely approaching the information theoretical limits at a moderate and manageable complexity has been the discovery of Turbo decoding for concatenated channel codes. The underlying Turbo principle of exchanging statistically independent extrinsic information between components has been extended to other concatenations of receiver elements.
In this thesis the concatenation of demodulation of higher order modulations and channel decoding, called bit-interleaved coded modulation with iterative decoding (BICM-ID), as well as the concatenation of channel decoding and soft decision source decoding, known as iterative source-channel decoding (ISCD), are investigated. To jointly exploit the inevitable redundancy in all three considered components in one Turbo process the novel Turbo DeCodulation (TDeC) scheme is developed. For all three Turbo systems manifold optimizations and designs are presented, including
* new mappings, signal constellation sets, index assignments, and quantizers;
* versatile highly redundant index assignments for exible multi-mode extensions;
* innovative channel code designs based on convolutional or block codes;
* novel methodologies for analysis, design, and optimization.
All systems are thoroughly analyzed by extrinsic information transfer (EXIT) charts and simulation results prove that the developed and appropriately designed systems exhibit a performance close to the information theoretical limits postulated by Shannon.

Download

BibTeX