Modeling extended L-band fiber amplifiers using neural networks trained on experimental data.

Producing high performance amplifiers requires accurate numerical models. As the optimization space is large, computationally efficient models are of great value. Parameter-based models for L-band amplifiers have accuracy limited by difficulty in estimating the Giles-parameter. The use a neural network model can avoid parametrization. We exploit a rich, experimentally captured training set to achieve a high accuracy neural network model. Our approach creates independent models for gain and noise figure. We examine both core and cladding pumping methods, again with independent models for each. The neural networks outperform parameter-based models with higher accuracy (variance of error reduced by 50%) and extremely fast simulation times (400 times faster), greatly facilitating amplifier design. As an example application, we design an amplifier to optimize optical signal-to-noise ratio by exhaustive search with our fast neural network models.

Fabrication of integrated waveguide grating in azo polymers.

In this Letter, we have introduced a technique, new to our knowledge, to fabricate gratings on a waveguide of azo-functionalized polymeric films using a slit mask and a fast, direct-writing method. To prevent the destruction of the waveguide by the grating formation on the waveguide, we placed a slit mask on the waveguide. By properly adjusting the resonance, this grating can be used as an integrated wavelength filter. We have produced an attenuation of 13.4 dB at 1562 nm with a FWHM of 3.45 nm. The grating has been fabricated as narrow as the width of the waveguide to couple filtered light into the waveguide by using a slit mask. Any light shifted from the resonance will pass through the waveguide undisturbed.