Enhanced Pub/Sub Communications for Massive IoT Traffic with SARSA Reinforcement Learning.
Carlos E. ArrudaPedro F. MoraesNazim AgoulmineJoberto S. B. MartinsPublished in: MLN (2020)
Keyphrases
- reinforcement learning
- reinforcement learning algorithms
- function approximation
- function approximators
- management system
- network traffic
- temporal difference learning
- state space
- machine learning
- markov decision processes
- model free
- action selection
- content based publish subscribe
- rl algorithms
- publish subscribe
- temporal difference
- real time
- eligibility traces
- traffic flow
- communication networks
- learning algorithm
- data analysis
- learning process
- dynamic programming
- network management
- road network
- big data
- optimal policy
- intrusion detection
- traffic control
- traffic congestion
- neural network
- cloud computing
- policy iteration
- single agent
- reward function
- transfer learning