An Offline Optical Path Planning Method Using Reinforcement Learning
Takafumi Tanaka, Katsuaki Higashimori · 2024
We proposed a new method for optimizing the optical path planning procedure by reinforcement learning for offline optical path planning problem. Simulations confirmed it achieved the frequency slot utilization efficiency comparable to Integer Linear Programming (ILP) for small networks and consistently outperforms various heuristic methods on real-scale networks.