Optimization and Improvement of BP Decoding Algorithm for Polar Codes Based on Deep Learning

Li Ge; Guiping Li

doi:10.2478/ijanmc-2023-0057

Figures & Tables

Multi-layer structure of deep neural network

Block diagram of neural network based decoder system

Performance of different network structures at N=8

Performance of different network structures at N=16

Performance of different network structures at N=32

Structure diagram of the proposed MLP-BP

Change of MLP-BP training loss value when N=128

Evolution of MLP-BP and BP BER when N=128

BER performance comparison of two decoding methods at N=32

BER performance comparison of two decoding methods at N=128

Decoding time delay

Algorithm	BP	MLP-BP
Decoding time delay	380	72

Proposed MLP-BP decoding algorithm

1: Enter. y₀, y₁, ⋯ y_N₋₁

2: Output. u₀, u₁, ⋯ u_N₋₁

3: Initialization: Initialization using (2) LLR(y_j)

4. for iter ←1 to iter_max do

5. for i ←n + 1 to n_NND do

6: Update using equation (3), Li,jiter

{\rm{L}}_{i,j}^{iter}

7. end for

8: After reaching NND use the sub-block NND_sub to calculate u_sub

9: u_sub After recoding to get x_sub

10: if after encoding x_sub by CRC checksum do

11: Using equation (7) yields, RnNND,subiter

R_{{n_{NND}},sub}^{iter}

12. end if

13: Retransmission

14. for i ← n_NND to n do

15: Update using equation (3)

Ri+1,jiter

R_{i + 1,j}^{iter}

16: end for

17: end for

Network Structure

	32-16-8	128-64-32	512-256-128
N=8	1024	11752	169846
N=16	1352	13488	174992
N=32	1352	13488	174992

Parameter Setting

Set options	Value
Test platform	Tensorflow
Encoding	Polar(32,16), (64,128)
Signal to noise ratio	1~5dB
loss function	Cross Entropy Loss
Optimizer	Adam

Parameters Settings

Parameters	Value
code length	8, 16, 32
code rate	0.5
batchsize	512
learning rate	0.001
training set size	106
epoch	103
network structure	32-16-8, 128-64-32, 512-256-128

Polar(32,16) Divided into Four Parts

Partitioning	Information bits	Relative Location	Code Rate
[0–7]	None	None	0
[8–15]	{11,12,13,14}	{3,5,6,7}	0.5
[16–23]	{19,21,22,23}	{3,5,6,7}	0.5
[24–31]	{24,25,26,27,28,29,30,31}	{0,1,2,3,4,5,6,7}	1

Polar(32,16) Divided into Two Parts

Partition	Information bits	Code Rate
[0–15]	{11,12,14,15}	0.25
[16–31]	{19,21,22,23,24,25,26,27,28,29,30,31}	0.75

Optimization and Improvement of BP Decoding Algorithm for Polar Codes Based on Deep Learning

Figures & Tables

Figure 1.

Figure 2.

Figure 3.

Figure 4.

Figure 5.

Figure 6.

Figure 7.

Figure 8.

Figure 9.

Figure 10.

Figure 11.

Figure 12.