Network-Wide SNR-based Channel Power Optimization

Spectral Power Profile Optimization of Field-Deployed WDM Network by Remote Link Modeling

“…In addition to the minimum margin gain, we also note the improvement of the margin in the long wavelength region by ≈ 2.2 dB and ≈ 0.8 dB for the long and short link, respectively. For comparison, the method from Garbhapu et al [15] achieved 0.5 dB margin improvement over a lab setup, where performance feedback from the receiver drives the power allocation. In this work, the performance improvement is available immediately and without requiring any feedback from the network.…”

“…We thereby demonstrate that an entire, deployed optical fiber network and its individual components can be modeled by only sending probe signals through a single access point on a single point to point link. The model is then used to optimize operation specific properties of the network [13]- [15] . In particular, the spectral power per WDM channel is optimized with a target of equalizing the signal to noise ratio (SNR) across all channels without requiring feedback on the quality of transmission as in Garbhapu et al [15] and without requiring gain flattening filters after the EDFAs [8] .…”

“…The model is then used to optimize operation specific properties of the network [13]- [15] . In particular, the spectral power per WDM channel is optimized with a target of equalizing the signal to noise ratio (SNR) across all channels without requiring feedback on the quality of transmission as in Garbhapu et al [15] and without requiring gain flattening filters after the EDFAs [8] .…”

Spectral power profile optimization of a field-deployed wavelength-division multiplexing network enabled by remote EDFA modeling

“…Joint optimization of input power and amplifier gain is shown in [42]. In [43,44] an heuristic algorithm is used to optimize SNR via the channel powers in a testbed with commercial, real-time equipment. Optimization of inline amplifier settings for a S+C+L multi-band system is demonstrated in [45].…”

“…The subsequent optimization of the launch power profiles shows improvement in optical signalto-noise ratio (OSNR)/SNR margin in a field deployed system compared to a flat launch power profile. As the system model is fully trained before the system optimization, our method does not require any feedback from the system as in [43], nor long system shutdown periods for characterization. The structure of this paper is as follows: Section II describes the models used to model fiber and EDFA devices, in particular we propose a novel hybrid EDFA model combining physical properties with an ML algorithm.…”