Laser & Optoelectronics Progress, Volume. 61, Issue 13, 1312001(2024)
Optical Performance Monitoring Based on Semi-Supervised Deep Learning
Fig. 2. ADTP collected without dispersion with OSNR for 16QAM at 16 dB, 23 dB, and 30 dB, and for 32QAM/64QAM at 21 dB, 28 dB, and 35 dB,respectively
Fig. 7. Relationship between dispersion and OSNR classification accuracy of different signals under three modulation formats obtained using different models at a data labeling rate of 10%. (a) 16QAM; (b) 32QAM; (c) 64QAM
Fig. 8. Relationship between dispersion and OSNR classification accuracy of different signals under three modulation formats obtained using different models at a data labeling rate of 5%. (a) 16QAM; (b) 32QAM; (c) 64QAM
Fig. 9. MAE monitoring error of 16QAM signal at different dispersion levels when data labeling rate is 5%
|
Get Citation
Copy Citation Text
Zhenwen Li, Xiyue Zhu, Yu Cheng. Optical Performance Monitoring Based on Semi-Supervised Deep Learning[J]. Laser & Optoelectronics Progress, 2024, 61(13): 1312001
Category: Instrumentation, Measurement and Metrology
Received: Jul. 10, 2023
Accepted: Nov. 8, 2023
Published Online: Jul. 17, 2024
The Author Email: Yu Cheng (chengyu@163.com)
CSTR:32186.14.LOP231679