在迭代解调中，demapper送给译码器的外信息为 P(c=bly) P(c=b.0) P(c=b:1)se Po.IIP(”=b,x) 它被译码器看作输入的先验信息。在对数域中，它可以表示为 Lwr(c0=b)=logP(c0=by,）-logP(c0=b） The SISO decoder for FEC code is used for generating the APPs P()of the information bits,.as well as the extrinsic APPs p(c:O)(应该是纯外信息，等于 APP(cIy)/P(c:),不包含信道值)of the coded bits,from the deinterleaved probabilities P(=).It can be implemented based on the MAP,Max-Log-MAPA and other such algorithms. Binary C, Bit interleaver Signal encode mapper Equivaler Channel P49,O) P(d:I P(v;O) Binary De-interleaver De- decoder mapper P(c:O) P(g:) 入 Interleaver Figure 5.11.2.BICM-ID scheme using soft-decision feedback ■Labeling method It should be pointed that labeling plays a key role in BICM.The performance of BICM depends on the labeling used.Fig.x shows the process of subset partitioning for each of the three bit positions for both Gray abelingand Splabeling that Gray labeling hasa ling isa more appropriate mapping during the first decoding iteration,and so it is adopted by the non-iterative BICM scheme of Fig.x.With iterative decoding and SP-based labeling. BICM-ID can achieve a high diversity order as well as a high FED. 23 23 在迭代解调中，demapper 送给译码器的外信息为 ( ) ( ) ( ) ( ) ( ) ( ) 1 ( ) ( ) ( ) 0 | ; ( | ) () ; i t b i m t t i j t t t t jt i x j t j i Pc b y P c bO P y x P c b x Pc bI − ∈ = ≠ ⎛ ⎞ = ⎜ ⎟ == = = = ⎝ ⎠ ∑ ∏ X 它被译码器看作输入的先验信息。在对数域中，它可以表示为 ( ) ( ) ( ) () () () log | log ; demapper i i i L c b Pc b y Pc bI et t t t == = − = The SISO decoder for FEC code is used for generating the APPs ( ) ( ); i Pu O t of the information bits, as well as the extrinsic APPs ( ) ( ); i Pc O t （应该是纯外信息，等于 ( ) ( | )/ ( ; ) i i APP c y P c I t t ，不包含信道值）of the coded bits, from the deinterleaved probabilities ( ) ( ) ; i P v bO t = . It can be implemented based on the MAP, Max-Log- MAPA and other such algorithms. Binary Bit interleaver encoder Signal mapper π b Equivalent Channel De￾mapper De-interleaver Binary decoder ut ct vt xt Interleaver yt ( ) ( ;) i Pv I t ( ) ( ;) i Pc O t ( ) ( ;) i Pv O t ( ) ( ;) i Pc I t ( ) ( ;) i Pu O t Figure 5.11.2. BICM-ID scheme using soft-decision feedback „ Labeling method It should be pointed that labeling plays a key role in BICM. The performance of BICM depends on the labeling used. Fig.x shows the process of subset partitioning for each of the three bit positions for both Gray labeling and SP labeling. It is seen that Gray labeling has a lower number of nearest neighbours compared to SP-based labeling. Hence, Gray labeling is a more appropriate mapping during the first decoding iteration, and so it is adopted by the non-iterative BICM scheme of Fig.x. With iterative decoding and SP-based labeling, BICM-ID can achieve a high diversity order as well as a high FED