Towards DRL-based Routing and Spectrum Assignment in Optical Networks: Lessons to be Learned from Markov Decision Processes.
Ronald Romero ReyesThomas BauschertPublished in: LATINCOM (2021)
Keyphrases
- markov decision processes
- optical networks
- wavelength division multiplexing
- link failure
- optimal policy
- service differentiation
- finite state
- state space
- decision theoretic planning
- policy iteration
- reinforcement learning
- transition matrices
- reachability analysis
- dynamic programming
- average cost
- routing and wavelength assignment
- average reward
- wdm networks
- infinite horizon
- ad hoc networks
- model based reinforcement learning
- partially observable
- planning under uncertainty
- state and action spaces
- routing protocol
- network topology
- action sets
- learning algorithm