TY - GEN
T1 - Extrinsic LLR Computation by the SISO LMMSE Detector: Four Different Approaches
AU - Haselmayr, Werner
AU - Springer, Andreas
PY - 2015/2
Y1 - 2015/2
N2 - In this paper we present four different approaches for the computation of the extrinsic information by a soft-input soft-output LMMSE detector used in a turbo equalization system. Moreover, we show the equivalence of the four approaches. All methods apply a Gaussian approximation, but at different stages, e.g. Gaussian approximation on the LMMSE filter output. Each approach offers different strategies for reducing the computational complexity, where the factor graph approach is the most promising solution.
AB - In this paper we present four different approaches for the computation of the extrinsic information by a soft-input soft-output LMMSE detector used in a turbo equalization system. Moreover, we show the equivalence of the four approaches. All methods apply a Gaussian approximation, but at different stages, e.g. Gaussian approximation on the LMMSE filter output. Each approach offers different strategies for reducing the computational complexity, where the factor graph approach is the most promising solution.
UR - https://www.scopus.com/pages/publications/84952334156
U2 - 10.1007/978-3-319-27340-2_66
DO - 10.1007/978-3-319-27340-2_66
M3 - Conference proceedings
SN - 9783319273396
T3 - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
SP - 529
EP - 536
BT - Proc. Int. Conf. Comput. Aided Syst. Theory 2015 (EUROCAST 2015)
A2 - Pichler, Franz
A2 - Moreno-Díaz, Roberto
A2 - Quesada-Arencibia, Alexis
ER -