DocumentCode :
1808562
Title :
Log domain implementation of the dual-APP algorithm
Author :
Srinivasan, Sudharshan ; Pietrobon, Steven S.
Author_Institution :
Inst. for Telecommun. Res., South Australia Univ., The Levels, SA
fYear :
2005
fDate :
2-4 Feb. 2005
Firstpage :
107
Lastpage :
111
Abstract :
For high rate codes, an a posteriori probability (APP) algorithm working on the dual code is preferable. The challenge involved in log domain implementation of this algorithm due to the presence of negative valued metrics is highlighted. To represent these metrics, a new representation system is proposed in place of the existing sign-magnitude representation. A generalised model of a trellis based implementation is presented, over which the log domain implementation of two approaches, one based on likelihood ratios (LRs) and the other based on reflection coefficients (RCs) are compared in complexity. It is found that the approach based on LRs is less complex
Keywords :
binary codes; computational complexity; convolutional codes; decoding; probability; a posteriori probability; binary convolutional code; dual-APP algorithm; likelihood ratios; log domain implementation; reflection coefficients; sign-magnitude representation; trellis based implementation; Arithmetic; Australia; Bandwidth; Convolutional codes; Differential equations; Iterative decoding; Lakes; Parity check codes; Reflection; Signal mapping; APP; Dual code; MAP decoding; SISO decoding;
fLanguage :
English
Publisher :
ieee
Conference_Titel :
Communications Theory Workshop, 2005. Proceedings. 6th Australian
Conference_Location :
Brisbane, Qld.
Print_ISBN :
0-7803-9007-5
Type :
conf
DOI :
10.1109/AUSCTW.2005.1624235
Filename :
1624235
Link To Document :
بازگشت