Digital Signal Processing Reference
In-Depth Information
confined to
M
and less side information make the decoding tree to have to prune
branches temporary with little Maximum A Posteriori Probability. Maybe, the
pruned branches are just in the proper path in the long run.
10
-1
10
-2
10
-3
10
-4
DOAC N=200
DOAC N=1000
DQAC N=200
DQAC N=1000
Proposed N=200
Propose
d
N=1000
10
-5
10
-6
0.1
0.15
0.2
0.25
0.3
0.35
0.4
0.45
H(X/Y)
p
=0.9 and
p
=0.1
Fig. 4.
BER vs. H(X/Y) for
0
1
Comparing Figure 3 and Figure 4, one can find, whatever length of the source
sequence is, and whatever the implement scheme is, the decoding performance for
asymmetric source sequence is better than that of symmetric source sequence. The
results show that DAC is more suited for asymmetric source sequences. This is an
advantage inheriting from the characteristic of AC.
4
Conclusions
DAC
is an effective implementation of Slepian-Wolf coding, especially for shorter
source sequence with asymmetric statistical characteristics. Two implementation
schemes of DAC in existing literatures is discussed. Then based on the advantage and
disadvantage of them, a new scheme of DAC is proposed. The proposed DAC
implements the encoding process using overlapped interval as that in DOAC to
produce FSM representation. Consequently, the decoding is implemented using
optimized List-Viterbi algorithm. The simulation results show that the proposed DAC
not only has consistent performance trend but is better than the two existing schemes.
Acknowledgement.
This work is acknowledged to Excellent Young Fund of
Shanghai Education Commission (sdl10003) and Shanghai Science and Technology
Commission Research Programs (10dz1501000).
