Search for: All records

We implement a cascaded learning framework leveraging three different EDFA and fiber component models for OSNR and GSNR prediction, achieving MAEs of 0.20 and 0.14 dB over a 5-span network under dynamic channel loading. 

Optical transmission systems require accurate modeling and performance estimation for autonomous adaption and reconfiguration. We present efficient and scalable machine learning (ML) methods for modeling optical networks at component- and network-level with minimized data collection. 

Scalable methods for optical transmission performance prediction using machine learning (ML) are studied in metro reconfigurable optical add-drop multiplexer (ROADM) networks. A cascaded learning framework is introduced to encompass the use of cascaded component models for end-to-end (E2E) optical path prediction augmented with different combinations of E2E performance data and models. Additional E2E optical path data and models are used to reduce the prediction error accumulation in the cascade. Off-line training (pre-trained prior to deployment) and transfer learning are used for component-level erbium-doped fiber amplifier (EDFA) gain models to ensure scalability. Considering channel power prediction, we show that the data collection process of the pre-trained EDFA model can be reduced to only 5% of the original training set using transfer learning. We evaluate the proposed method under three different topologies with field deployed fibers and achieve a mean absolute error of 0.16 dB with a single (one-shot) E2E measurement on the deployed 6-span system with 12 EDFAs. 

We implement a cascaded learning framework using component-level EDFA models for optical power spectrum prediction in multi-span networks, achieving a mean absolute error of 0.17 dB across 6 spans and 12 EDFAs with only one-shot measurement. 

We report the first field demonstration of 4D link tomography using a commercial transponder, which offers distance, time, frequency, and polarization-resolved monitoring. This scheme enables autonomous transponders that identify locations of multiple QoT degradation causes. 

We proposed the use of BOTDA as a monitoring tool to identify fiber types present in deployed hybrid-span fiber cables, to assist in network planning, setting optimal launch powers, and selecting correct modulation formats. 

Abstract Litter decomposition is a key ecological process that determines carbon (C) and nutrient cycling in terrestrial ecosystems. The initial concentrations of C and nutrients in litter play a critical role in this process, yet the global patterns of litter initial concentrations of C, nitrogen (N) and phosphorus (P) are poorly understood.We employed machine learning with a global database to quantitatively assess the global patterns and drivers of leaf litter initial C, N and P concentrations, as well as their returning amounts (i.e. amounts returned to soils).The medians of litter C, N and P concentrations were 46.7, 1.1, and 0.1%, respectively, and the medians of litter C, N and P returning amounts were 1.436, 0.038 and 0.004 Mg ha⁻¹ year⁻¹, respectively. Soil and climate emerged as the key predictors of leaf litter C, N and P concentrations. Predicted global maps showed that leaf litter N and P concentrations decreased with latitude, while C concentration exhibited an opposite pattern. Additionally, the returning amounts of leaf litter C, N and P all declined from the equator to the poles in both hemispheres.Synthesis: Our results provide a quantitative assessment of the global concentrations and returning amounts of leaf litter C, N and P, which showed new light on the role of leaf litter in global C and nutrients cycling.