S. G. Samúelsson and E. Hyytiä, Applying Reinforcement Learning to Basic Routing Problem, in 13th International Conference on Queueing Theory and Network Applications (QTNA2018), Springer, Lecture Notes in Computer Science, vol 10932, 2018, Tsukuba, Japan.
Abstract: Routing jobs to parallel servers is a common and important task in today's computer and communication systems. As each routing decision affects the jobs arriving later, and determining the (near) optimal decisions is non-trivial. In this paper, we apply reinforcement learning techniques to the elementary job routing problem with heterogeneous servers and a general cost structure. We study the convergence of the reinforcement learning to a near-optimal policy (that we can determine by other means), and compare its performance against heuristic routing policies such as Join-the-Shortest-Queue (JSQ) and Shortest-Expected-Delay (SED).
Links:
BibTeX entry:
@inproceedings{samuelsson-qtna-2018,
title = {Applying Reinforcement Learning to Basic Routing Problem},
author = {Sigur{\dh}ur Gauti Sam{\'u}elsson and Esa Hyyti{\"a}},
booktitle = {13th International Conference on Queueing Theory and Network Applications ({QTNA2018})},
month = {Jul.},
year = {2018},
doiopt = {10.1007/978-3-319-93736-6_18},
address = {Tsukuba, Japan},
editor = {Takahashi Y. and Phung-Duc T. and Wittevrongel S. and Yue W.},
series = {Lecture Notes in Computer Science, vol 10932},
publisher = {Springer},
}