Multi-Context-aware RL approach towards INT-based Congestion Control Algorithm
Ramyashree Venkatesh Bhat, Jetmir Haxhibeqiri, Ingrid Moerman, Jeroen Hoebeke · 2024
Diverse applications in private networks need to be optimally served over both wired and wireless. One way to do so is by having more intelligent protocols, as current designs only have very limited feedback. This paper proposes a reinforcement learning-based congestion control algorithm for private wireless networks that takes into account application information and network context. Three algorithm designs are presented in the paper, and one was implemented and tested, showing adaptability to changing contexts.