Deep Learning Aided Channel Estimation and Signal Detection for Underwater Optical Communication

Parameters	Value
Transmission distance under water type 1 (coastal) /m	15
Transmission distance under water type 2 (clear) /m	20
Number of OOFDM symbols in one frame	5
DC-bias of DCO-OFDM	0.18
Number of OOFDM subcarriers	192
Modulation constellation	16QAM

Table 2. UWOC channel parameters used in simulation

View table

Table 2. UWOC channel parameters used in simulation

Parameters	Value
Number of water types for training	2
Training water type 1 (coastal), {c₁，a₁，b₁}	{0.179，0.219，0.398}
Training water type 2 (clear), {c₂，a₂，b₂}	{0.114，0.037，0.151}
Test water type 1 (coastal), {c₁，a₁，b₁}	{0.179，0.219，0.398}
Test water type 2 (clear), {c₂，a₂，b₂}	{0.114，0.037，0.151}

Table 3. Complexity comparison of various schemes
View table
Table 3. Complexity comparison of various schemes
Scheme FLOPs per OOFDM symbol
LS 4.224×10³
MMSE 4.80×10⁶
DNN 2.56×10³

Tools

Get Citation

Copy Citation Text

Jia Shi, Aiping Huang, Linwei Tao. Deep Learning Aided Channel Estimation and Signal Detection for Underwater Optical Communication[J]. Chinese Journal of Lasers, 2022, 49(17): 1706004

Download Citation

EndNote(RIS)BibTex Plain Text

Set citation alerts for article

Save article for my favorites