TCP-Drinc: Smart Congestion Control Based on Deep Reinforcement Learning

Xavier : A Reinforcement-Learning Approach to TCP Congestion Control

Controlling congestion is a fundmanetal problem in computer networks. If the input load is greater than the output bandwidth at a particular switch, the bottleneck’s queue begins to fill up and we say that it is congested. In pathological scenarios and under certain protocols, the saturation of buffers, or bufferbloat [5], can lead to congestion collapse, a condition in which congestion reaches...

Operation Scheduling of MGs Based on Deep Reinforcement Learning Algorithm

: In this paper, the operation scheduling of Microgrids (MGs), including Distributed Energy Resources (DERs) and Energy Storage Systems (ESSs), is proposed using a Deep Reinforcement Learning (DRL) based approach. Due to the dynamic characteristic of the problem, it firstly is formulated as a Markov Decision Process (MDP). Next, Deep Deterministic Policy Gradient (DDPG) algorithm is presented t...

On Hybrid TCP Congestion Control

This paper presents several considerations on the hybrid TCP congestion control, in which lossbased schemes like TCP-Reno and delay-based schemes like TCP-Vegas are combined in a smart way. We firstly introduce simple analytical models of loss-based TCP, delay-based TCP and their hybrid. With support of experimental results, effectiveness of the hybrid TCP is proved. Then, we try to develop a m...

Dynamic Generalization Kanerva Coding in Reinforcement Learning for TCP Congestion Control Design

Traditional reinforcement learning (RL) techniques often encounter limitations when solving large or continuous stateaction spaces. Training times needed to explore the very large space are impractically long, and it can be difficult to generalize learned knowledge. A compact representation of the state space is usually generated to solve both problems. However, simple state abstraction often c...

exploring delay - based tcp congestion control