@INPROCEEDINGS{Espo2110:Congestion,
AUTHOR="Hunter M Park and Flavio Esposito",
TITLE="{Congestion-Aware} Routing via Reinforcement Learning with Graph
Convolutional Networks",
BOOKTITLE="2021 12th International Conference on Network of the Future (NoF) (NoF
2021)",
ADDRESS="University of Coimbra, Portugal",
DAYS=5,
MONTH=oct,
YEAR=2021,
KEYWORDS="routing; reinforcement learning; Graph Convolutional Networks; congestion
control",
ABSTRACT="The computer networking community has been steadily increasing
investigations into the use of machine learning to help solve tasks such as
routing, traffic prediction, and resource management. The traditional
best-effort nature of Internet connections allows a single link to be
shared among multiple flows competing for network resources, often without
consideration of in-network states. In particular, due to the recent
successes in other applications, Reinforcement Learning has seen steady
growth in network management, in general, and routing in particular.
If there are changes in the network topology, however, retraining is often
required to avoid significant performance losses. This restriction has
mostly prevented the deployment of Reinforcement Learning-based routing in
real environments. In this paper, we approach routing as a reinforcement
learning problem with two novel twists: minimize flow set collisions, and
construct the reinforcement learning policy such that it is capable of
routing in dynamic network conditions without the need for retraining. We
then compare this approach to other routing protocols, with respect to a
variety of Quality-of-Service metrics. Lastly, we discuss directions for
continued research into Reinforcement Learning-based routing."
}