Iterative decoding for rate adaptive forward error correction on the mobile satellite channel.

Abstract

In this thesis, we examine rate adaptive forward error correction codes for the mobile satellite channel. We are especially interested in codes that can be iteratively decoded. Multilevel coded modulations are considered: partitioning of constellations, coding principle and iterative decoding are studied. We present design rules for multilevel coded modulations through the asymptotic coding gain, the minimum Hamming distance of the code and with information theory arguments. Computer simulations have been run to confirm the validity of the design rules and to determine the impact of the interleaver size on the performance of a coded modulation scheme. A new construction for multilevel coded modulation with parallel concatenation is suggested. Computer simulations have confirmed the efficiency of this new technique. Product codes are also considered: code construction and iterative decoding algorithms are studied. We focus on the two-dimensional (24, 12) Golay product code and find its performances on AWGN and Rayleigh fading channels through computer simulations. Three rate coding systems for the mobile satellite channel are presented. A performance criterion for speech transmission with such systems have been determined. Adaptive rate techniques can save up to 5.0 dB and 8.8 dB (depending on the type of environment) in terms of transmitted power.