Fig. 4: Experimental setup.
From: Deep learning prediction of noise-driven nonlinear instabilities in fibre optics
Schematic of the experimental setup implemented for reconfigurable modulation instability seeding and real-time spectral characterization. The data obtained from dispersive Fourier transform (DFT) signal processing and real-time optoelectronic acquisition can be leveraged for artificial neural network training after suitable statistical analysis. (EDFA: Erbium doped fibre amplifier; ASE: amplified spontaneous emission; HNLF: highly-nonlinear fibre; DCF: dispersion compensating fibre; OSA: optical spectrum analyzer).
