Abstract:The breathtaking progress in machine learning has motivated studies in learning-based congestion control algorithms, which are expected to adjust congestion window choices according to the dynamic network environment. Nevertheless, existing learning-based congestion control protocols are mostly designed and trained for a specific network environment. When applied to a different network environment, the previously trained model may see a considerable degradation in performance. As the rapid development of communication technologies has given rise to the emergence of a diversity of new networks, it is desirable for learning-based congestion control models to quickly transfer to different network environments. Driven by this motivation, we propose a novel transferable congestion control (TRCC) protocol, which takes full advantage of both reinforcement learning (RL) and transfer learning to intelligently cope with network congestion in scenarios. The key idea to enable a fast transfer from the source network environment to the target network environment is to fine-tune a well-trained model in the source network to suit the target network in a time-effective way. We theoretically prove the transferability and quick convergence of our proposed transfer RL-based congestion control algorithm by deriving the Markov transition matrix and the similarity of the reward function. Our experiments validate that TRCC can converge in a new network environment in a short time while achieving comparable performance with baseline algorithms.

Design Of Atm Congestion Controller Based On Reinforcement Learning Method

TRCC: Transferable Congestion Control with Reinforcement Learning

An Active Queue Management Scheme Using Neural Network Based Predictive Control

Reinforcement Learning Approach to ABR Traffic Control of ATM Networks

Simulated Annealing-Reinforcement Learning Algorithm For Abr Traffic Control Of Atm Networks

Simulated Annealing Q-learning Algorithm for ABR Traffic Control of ATM Networks

Design of ATM Networks Congestion Controller Based on the Virtual Queue

Congestion Control Scheme in Atm Networks Based on Fast Tracing-Queue

Research on Congestion Control of ATM Networks Based on Interior Model Control

RLCC: Practical Learning-based Congestion Control for the Internet

Integral Sliding Mode Controller Design for Congestion Problem in ATM Networks

Congestion Control of ATM Networks with Multiple Bottlenecks

A Novel Algorithm of Robust Congestion Control in Multi-Virtual-circuit ATM Networks

Reinforcement Learning Based Congestion Control in a Real Environment.

Robust Controller Design For Atm Network With Time-Varying Multiple Time-Delays

Ftqcs: Fast Tracing-Queue Control Scheme In Atm Networks

Congestion Control Based on Predictive Control in ATM Networks

Optimal Control Theory Based ABR Traffic Congestion Control in ATM Network with Multiple Congested Nodes

Binary Abr Congestion Control Based On Discrete-Time Variable Structure Control Algorithm

A Distributed Reinforcement Learning Approach to In-network Congestion Control

A Q-learning Model-Independent Flow Controller for High-Speed Networks