Theoretical and artificial neural network modeling for the output power of irradiated erbium doped fiber amplifier

The out power of the erbium doped fiber amplifiers (EDFAs) irradiated by neutrons with a dose of 1000 krad, it is investigated by theoretical and artificial neural network (ANN) modeling. Levenberg–Marquardt algorithm has been used for the ANN modeling. The theoretical as well as ANN models are in good agreement with the experimental data. A minimum acceptable error of 1.4×10−4 has been obtained. The theoretical and the ANN models help us to understand the output of EDFAs in detail. The EDF could be used only in the low radiation dose less than 20 krad. The ANN model provide a powerful tool to study the output power of EDFAs.

► Theoretical and artificial neural networks models have been used to investigate the out power of EDFA.
► Training for experimental data using neural networks.
► The ANN model provide a powerful tool to study the EDFA output.