We could have maintained a list of such trellis paths where they
would be ranked in some order. This type of decoding is called list decoding.
Soft Decision Decoding. In this section we will present a decoder that allows us to distinguish
between small amplitude bits and large amplitude bits. In other words, we wish to reduce the influence
of questionable bits on the decoding procedure. Here bits that are more reliable impact the
decoding decision more than those that are unreliable (or less reliable). This can be accomplished
with the help of the following block diagram (see Fig. 5.35).
PERFORMANCE IMPROVEMENT TECHNIQUES 249
a
b
c
d
0
1
2
0
2
0
1
(r = 00)
0
2
2
1
1
1
1
0
a
b
c
d
0
0
2
0
2
1
1
0
2
0
3
3
(r = 00)
FIGURE 5.34 Survivor state transitions of the Viterbi decoding algorithm for r 11 01 01 00.
Demodulator
Viterbi
Algorithm
Received
Symbols
Decoded Bit
Stream
Multilevel Quantizer
FIGURE 5.35 Soft decision decoding overview.
Here the demodulator output amplitudes are preserved with the use of the multilevel quantizer.
When the bit amplitudes are preserved prior to entering the FEC decoder, this is called soft decision
decoding. Because of this multiple-quantized bit representation of a code word, the Hamming distance
no longer applies as a useful branch metric to us.
Pages:
390
391
392
393
394
395
396
397
398
399
400
401
402
403
404
405
406
407
408
409
410
411
412
413
414