LLM-enabled Full-stack Configuration Automation of SDM Transport Network
Conference: ECOC 2024 - 50th European Conference on Optical Communication
09/22/2024 - 09/26/2024 at Frankfurt, Germany
Proceedings: ITG-Fb. 317: ECOC 2024
Pages: 4Language: englishTyp: PDF
Authors:
Wang, Cen; Wakayama, Yuta; Yoshikane, Noboru; Tsuritani, Takehiro
Abstract:
Through leveraging language understanding power of GPT4.0 and GPT3.5-turbo, we design, evaluate and experimentally validate a full-stack configuration automation framework on a deployed SDM network including configuration generation and configuration error analysis tasks with 100 and 91.05 scores, respectively.