The Optical RL-Gym: An open-source toolkit for applying reinforcement learning in optical networks.
Carlos NatalinoPaolo MontiPublished in: ICTON (2020)
Keyphrases
- optical networks
- reinforcement learning
- open source
- wavelength division multiplexing
- function approximation
- model free
- service differentiation
- markov decision processes
- reinforcement learning algorithms
- state space
- rl algorithms
- optimal policy
- multi agent
- temporal difference
- learning algorithm
- optimal control
- source code
- wdm networks
- routing and wavelength assignment
- dynamic programming
- autonomous learning
- real time
- network technologies
- direct policy search
- learning agents
- control policy
- action selection
- reinforcement learning methods
- actor critic
- action space
- learning problems
- partially observable domains
- general purpose
- case study
- web services