No abstract

No abstract

We propose a machine learning based method to estimate the proportion of inter-channel and intra-channel nonlinear effects. The model is tested with simulations over different power profiles, fiber attenuations, span lengths and chromatic dispersions.

Submarine systems have recently evolved from turnkey systems into an open cable approach, where new metrics describing wet plant performance have been defined. The transmission GSNR has been standardized and is measured together with the OSNR in cable commissioning to characterize open submarine links. We propose in this paper a method based on numerical simulation to accurately predict the achievable capacity of open cables using only the commissioning parameters. We also assess the impact of the measurement uncertainties during commissioning on capacity prediction. Finally, we apply the proposed method to realistic subsea links and show how the uncertainty on the capacity estimate can be reduced further when using the commissioning measurements to reduce the uncertainty on line parameters.