Wolfgang Schairer

Dec 2, 2024 | opg.optica.org | Jaroslaw E. Prilepsky |Pedro Freire |Antonio Napoli |Wolfgang Schairer

We address the development of efficient neural network (NN)-basedpost-equalizers in long-haul coherent-detection densewavelength-division multiplexing (DWDM) optical transmission systems. To achieve a high level of generalization of the NN-based equalizers,we propose to employ multi-task learning (MTL). MTL refers to a singleshared machine learning (NN) model that can perform multiple different(albeit related) tasks.