Investigation on Normalized Factor and Finite Word-Length for 802.11n LDPC Decoding

in this paper, we theoretically derive the suitable normalized factor of normalized BP-based algorithm of LDPC decoding in 802.11n system. To be specific, we compute 1.15, 1.25, 1.3, 1.35 as the suitable normalized factors with regard to R=1/2, R=2/3, R=3/4, R=5/6 code in 802.11n system respectively. These factors can compensate for the performance loss in UMP BP-based algorithm. By analyzing the statistical distribution of variable node messages (R=1/2 code particularly), the word-length of variable nodes is reduced to 7 bits, and the simulation results show that it has a similar decoding performance compared with the theoretical 9-bit word-length decoding.