Maximizing Fiber Cable Capacity Under A Supply Power Constraint Using Deep Neural Networks

Abstract

We experimentally achieve a 19% capacity gain per Watt of electrical supply power in a 12-span link by eliminating gain flattening filters and optimizing launch powers using deep neural networks in a parallel fiber context.

Power Evolution Modeling and Optimization of Fiber Optic Communication Systems With EDFA Repeaters

“…4 is trained to predict the PSD output of the total system, similar to e.g. [10], [11]. This model is expected to provide an improved accuracy.…”

“…Some solutions exist for countering the SRS effect by launch PSD optimization [7], [8], [9], which do not take into account the gain tilt of practical EDFAs without a GFF. Alternatively, ML can be applied for the end-to-end system [3], [10], [11], which at least in the linear region of transmission can be employed for PSD optimization. Such solutions require a lot of training data to be generated for each link in the network, which is time consuming, and is furthermore susceptible to even minor changes in the link.…”

Modeling EDFA Gain: Approaches and Challenges

“…Some ASE noise models can be derived using schemes that are similar to gain modeling schemes [16,17]. Models that simultaneously calculate the signal power and ASE noise power have also been proposed [18,19].…”

“…ML-based modeling schemes to accurately estimate both the gain and the ASE noise. Some ML-based models can be used to estimate the signal power and the noise power of EDFA simultaneously [18,19]. Currently, simple NNs are utilized for these models, and a large dataset is needed.…”