Title :
LT coded MSK over AWGN channels
Author :
Hussain, Iqbal ; Xiao, Ming ; Rasmussen, Lars K.
Author_Institution :
Commun. Theor. Lab., KTH - R. Inst. of Technol., Sweden
Abstract :
We investigate the design of Luby Transform (LT) codes with minimum-shift-keying (MSK) modulation over additive white Gaussian noise (AWGN) channels. Both systematic and nonsystematic LT codes are considered from the perspectives of decoding threshold and bit error rate. Using systematic LT codes, coding complexity can be reduced by eliminating the need for decoder-doping, and less parity bits are required as compared to the nonsystematic LT codes. In addition, systematic LT codes exhibit better performance than their nonsystematic counterparts for low fixed code rates. To evaluate the performance of the LT coded MSK system, we consider a family of rate-compatible fixed-rate codes, which allows us to use extrinsic information transfer (EXIT) charts for analysis and design. We observe that systematic LT codes have a better decoding threshold than their corresponding nonsystematic counterparts for all tested rates up to R = 0.59 bits per symbol. Furthermore, with the use of EXIT charts, we propose a new degree distribution function for the LT code, which leads to improved decoding thresholds for all examined code rates, as compared to codes using existing degree distributions optimized for binary erasure and AWGN channels. Numerical results confirm our EXIT chart analysis and conclusions.
Keywords :
AWGN channels; binary codes; communication complexity; decoding; error statistics; minimum shift keying; transform coding; AWGN channel; EXIT chart analysis; LT coded MSK system; Luby transform code; additive white Gaussian noise channel; binary erasure channel; bit error rate; coding complexity; decoder-doping; decoding; degree distribution function; extrinsic information transfer chart; minimum-shift-keying modulation; nonsystematic LT code; parity bits; rate-compatible fixed-rate code; systematic LT code; Artificial neural networks; Lead;
Conference_Titel :
Turbo Codes and Iterative Information Processing (ISTC), 2010 6th International Symposium on
Conference_Location :
Brest
Print_ISBN :
978-1-4244-6744-0
Electronic_ISBN :
978-1-4244-6745-7
DOI :
10.1109/ISTC.2010.5613883