Core Concepts
Integrating physical-parameter-aware technologies for semi-automatic line-system provisioning optimizes transmission performance.
Abstract
This research article focuses on proposing methods and architecture for semi-automatic line-system provisioning using integrated physics-aware technologies. The study demonstrates the optimization of optical fiber line systems through digital longitudinal monitoring (DLM) and optical line system (OLS) physical parameter calibration. The methodology offers advantages such as minimized footprint, accurate estimation of network characteristics, and remote operation capability. Field trials at Duke University successfully showcased 1-hour provisioning with reduced QoT prediction errors compared to existing designs.
Directory:
- Introduction
- Optical networks expansion due to technology adoption.
- Related Works
- Challenges in commercial use of machine learning.
- Methodology
- Integration of DLM and OLS calibration for parameter extraction.
- Control Architecture
- Hybrid controller architecture for remote operation.
- Experimental Setup
- Field trial setup at Duke University with detailed fiber routes.
- Results: 1-Hour Provisioning & Maintenance
- Execution time breakdown and physical parameter visualization.
- Future Challenges
- Addressing scalability, automation feasibility, and fault detection.
- Conclusion
- Summary of proposed approach benefits and future research directions.
Stats
"We confirmed that our methodology has a reduction in the QoT prediction errors (±0.3 dB) over existing design (±0.6 dB)."
"The RMS error between OTDR traces and DLM was 0.45 dB."
"The average noise figure value for EDFAs in both ALLs is assumed 6.5 dB."