RSCAT: Towards zero touch congestion control based on actor-critic reinforcement learning and software-defined networking.
Gustavo DielCharles Christian MiersMaurício Aronne PillonGuilherme Piêgas KoslovskiPublished in: J. Netw. Comput. Appl. (2023)
Keyphrases
- actor critic
- reinforcement learning
- congestion control
- end to end
- temporal difference
- policy gradient
- optimal control
- reinforcement learning algorithms
- function approximation
- approximate dynamic programming
- gradient method
- neuro fuzzy
- packet loss
- average reward
- policy iteration
- peer to peer
- markov decision processes
- model free
- dynamic programming
- quality of service
- state space
- learning algorithm
- network bandwidth
- optimal policy
- machine learning
- window size
- step size
- learning problems
- learning tasks
- recursive least squares
- cost function
- multi agent