| | |
Summary: Turbo Codes for Binary Markov Sources 1
Guang-Chong Zhu and Fady Alajaji
Dept. of Mathematics and Statistics
Queen's University, Kingston
Ontario, Canada, K7L 3N6
E-mail: fzhugc, fadyg@mast.queensu.ca
Abstract | The reliable transmission via Turbo
codes of binary stationary ergodic Markov sources
over noisy channels is investigated. The rst con-
stituent Turbo decoder is designed to exploit the
source redundancy according to a modi ed ver-
sion of Berrou's original decoding method that
employs the Gaussian assumption on the extrin-
sic information. Due to interleaving, the sec-
ond constituent decoder is unable to adopt the
same decoding method. However, its extrinsic in-
formation is modi ed via a weighted correction
term. Substantial gains (from 1.29 up to 3.57 dB)
over the original Turbo codes are demonstrated.
Furthermore, the gaps to the Shannon limit are
|
