Adaptação de técnica de interpolação linear para a estimação do sinal de saída de amplificadores ópticos com entrada não plana

Abstract

The Internet has been the scene of increased data traffic. Given this consumption, optical communication networks are a good option to use because of their large bandwidth. In them, optical amplifiers play a fundamental role in compensating for power losses that the signal suffers during transmission, while degrading it through the addition of distortions and noise. It is necessary to adapt them so that their operation is autonomous and favors transmission quality. Thus, the ACOP (Adaptive Control of Operating Point) problem arises, whose objective is to adapt the amplifier configuration according to the state of the network. In the solutions in the literature, it is essential that the set of amplifier operating points, called the power mask, is known. To offer greater granularity in the space of the operating points defined in the masks, there are techniques capable of estimating the amplifier’s behavior at undefined points, but not all of them take into account non-linearities in the signals, common in real-world scenarios. This work proposes an adaptation of one of the optical signal estimation techniques in the literature, based on linear interpolation, in order to make it consider power masks with non-plane signals. The evaluations show that the adaptation was able to improve the performance of the original technique, which presented error measures much greater than those of the technique proposed here. Furthermore, the adaptation was able to obtain similar results and, in several cases, better than another technique in the literature, based on artificial neural networks.