Optimization of four-wave mixing wavelength conversion in a quantum-dot semiconductor optical amplifier based on the genetic algorithm

A novel approach based on the artificial neural network (ANN) and the genetic algorithm (GA) is presented for optimization of four-wave mixing (FWM) wavelength conversion in a quantum dot semiconductor optical amplifier (QD-SOA). First of all, we propose a simple, accurate, and fast model based on the feedforward ANN for the characteristics of FWM in a QD-SOA. To train the ANN, we collect the required data from a numerical model. In this model, the efficiency of FWM is obtained numerically taken into account the effect of pump/probe and the occupation probability of energy levels by using the slice technique. Then, the optimal design of QD-SOA as the FWM wavelength converter is performed using the GA.