Prediction of osnr in a 7-channel dwdm/udwdm transmission system under the combined effects of FWM, SRS and ASE

Abstract

In this study, an artificial neural network (ANN) model has been developed to predict the optical signal-to-noise ratio (OSNR) at the center channel of a 7-channel DWDM/UDWDM transmission system using an erbium-doped fiber amplifier (EDFA), taking into account the triple effects of four-wave mixing (FWM), stimulated Raman scattering (SRS), and amplified spontaneous emission (ASE) noise. The dataset used to predict the OSNR value contains 1200 distinct data and it has been created considering a 7-channel DWDM/UDWDM transmission system employing 1, 2, 4, and 5 EDFAs with channel spacing values of 3.125 GHz, 6.25 GHz, 12.5 GHz, 25 GHz, 50 GHz, and 100 GHz. Additionally, it has been assumed that the channel input power varied within the range from 0.1 mW to 5 mW. The obtained data has been divided into 70% for training, 15% for validation, and 15% for testing.In the study, the mean absolute error (MAE) has been calculated as 0.7935, the mean absolute percentage error (MAPE) has been computed as 9.3737, the mean squared error (MSE) has been obtained as 1.1487, and the root mean squared error (RMSE) has been found as 1.0718.