Optical SNR estimation using machine learning
Keyword(s):
Abstract A technique for the estimation of an optical signal-to-noise ratio (OSNR) using machine learning algorithms has been proposed. The algorithms are trained with parameters derived from eye-diagram via simulation in 10 Gb/s On-Off Keying (OOK) nonreturn-to-zero (NRZ) data signal. The performance of different machine learning (ML) techniques namely, multiple linear regression, random forest, and K-nearest neighbor (K-NN) for OSNR estimation in terms of mean square error and R-squared value has been compared. The proposed methods may be useful for intelligent signal analysis in a test instrument and to monitor optical performance.
2020 ◽
Vol 18
(1)
◽
pp. 200
2020 ◽
Vol 11
(9)
◽
pp. 442-449
◽
2021 ◽
Vol 10
(3)
◽
pp. 14-25
2020 ◽
2020 ◽
Vol 8
(2S12)
◽
pp. 65-70