Matches in SemOpenAlex for { <https://semopenalex.org/work/W4311117974> ?p ?o ?g. }
Showing items 1 to 79 of
79
with 100 items per page.
- W4311117974 abstract "We approach the task of network congestion control in datacenters using Reinforcement Learning (RL). Successful congestion control algorithms can dramatically improve latency and overall network throughput. Until today, no such learning-based algorithms have shown practical potential in this domain. Evidently, the most popular recent deployments rely on rule-based heuristics that are tested on a predetermined set of benchmarks. Consequently, these heuristics do not generalize well to newly-seen scenarios. Contrarily, we devise an RL-based algorithm with the aim of generalizing to different configurations of real-world datacenter networks. We overcome challenges such as partial-observability, non-stationarity, and multi-objectiveness. We further propose a policy gradient algorithm that leverages the analytical structure of the reward function to approximate its derivative and improve stability. We show that this scheme outperforms alternative popular RL approaches, and generalizes to scenarios that were not seen during training. Our experiments, conducted on a realistic simulator that emulates communication networks' behavior, exhibit improved performance concurrently on the multiple considered metrics compared to the popular algorithms deployed today in real datacenters. Our algorithm is being productized to replace heuristics in some of the largest datacenters in the world." @default.
- W4311117974 created "2022-12-23" @default.
- W4311117974 creator A5002549620 @default.
- W4311117974 creator A5010495410 @default.
- W4311117974 creator A5011863840 @default.
- W4311117974 creator A5021079458 @default.
- W4311117974 creator A5036260775 @default.
- W4311117974 creator A5045719865 @default.
- W4311117974 creator A5059390892 @default.
- W4311117974 creator A5088301119 @default.
- W4311117974 date "2021-02-18" @default.
- W4311117974 modified "2023-09-27" @default.
- W4311117974 title "Reinforcement Learning for Datacenter Congestion Control" @default.
- W4311117974 doi "https://doi.org/10.48550/arxiv.2102.09337" @default.
- W4311117974 hasPublicationYear "2021" @default.
- W4311117974 type Work @default.
- W4311117974 citedByCount "0" @default.
- W4311117974 crossrefType "posted-content" @default.
- W4311117974 hasAuthorship W4311117974A5002549620 @default.
- W4311117974 hasAuthorship W4311117974A5010495410 @default.
- W4311117974 hasAuthorship W4311117974A5011863840 @default.
- W4311117974 hasAuthorship W4311117974A5021079458 @default.
- W4311117974 hasAuthorship W4311117974A5036260775 @default.
- W4311117974 hasAuthorship W4311117974A5045719865 @default.
- W4311117974 hasAuthorship W4311117974A5059390892 @default.
- W4311117974 hasAuthorship W4311117974A5088301119 @default.
- W4311117974 hasBestOaLocation W43111179741 @default.
- W4311117974 hasConcept C111919701 @default.
- W4311117974 hasConcept C112972136 @default.
- W4311117974 hasConcept C119857082 @default.
- W4311117974 hasConcept C120314980 @default.
- W4311117974 hasConcept C127705205 @default.
- W4311117974 hasConcept C154945302 @default.
- W4311117974 hasConcept C157764524 @default.
- W4311117974 hasConcept C158379750 @default.
- W4311117974 hasConcept C195563490 @default.
- W4311117974 hasConcept C28826006 @default.
- W4311117974 hasConcept C31258907 @default.
- W4311117974 hasConcept C33923547 @default.
- W4311117974 hasConcept C36299963 @default.
- W4311117974 hasConcept C41008148 @default.
- W4311117974 hasConcept C555944384 @default.
- W4311117974 hasConcept C76155785 @default.
- W4311117974 hasConcept C82876162 @default.
- W4311117974 hasConcept C97541855 @default.
- W4311117974 hasConceptScore W4311117974C111919701 @default.
- W4311117974 hasConceptScore W4311117974C112972136 @default.
- W4311117974 hasConceptScore W4311117974C119857082 @default.
- W4311117974 hasConceptScore W4311117974C120314980 @default.
- W4311117974 hasConceptScore W4311117974C127705205 @default.
- W4311117974 hasConceptScore W4311117974C154945302 @default.
- W4311117974 hasConceptScore W4311117974C157764524 @default.
- W4311117974 hasConceptScore W4311117974C158379750 @default.
- W4311117974 hasConceptScore W4311117974C195563490 @default.
- W4311117974 hasConceptScore W4311117974C28826006 @default.
- W4311117974 hasConceptScore W4311117974C31258907 @default.
- W4311117974 hasConceptScore W4311117974C33923547 @default.
- W4311117974 hasConceptScore W4311117974C36299963 @default.
- W4311117974 hasConceptScore W4311117974C41008148 @default.
- W4311117974 hasConceptScore W4311117974C555944384 @default.
- W4311117974 hasConceptScore W4311117974C76155785 @default.
- W4311117974 hasConceptScore W4311117974C82876162 @default.
- W4311117974 hasConceptScore W4311117974C97541855 @default.
- W4311117974 hasLocation W43111179741 @default.
- W4311117974 hasOpenAccess W4311117974 @default.
- W4311117974 hasPrimaryLocation W43111179741 @default.
- W4311117974 hasRelatedWork W14296039 @default.
- W4311117974 hasRelatedWork W1486948142 @default.
- W4311117974 hasRelatedWork W1566886392 @default.
- W4311117974 hasRelatedWork W1993378674 @default.
- W4311117974 hasRelatedWork W2298102683 @default.
- W4311117974 hasRelatedWork W2385763152 @default.
- W4311117974 hasRelatedWork W3133382421 @default.
- W4311117974 hasRelatedWork W3161650376 @default.
- W4311117974 hasRelatedWork W4283803449 @default.
- W4311117974 hasRelatedWork W4284897198 @default.
- W4311117974 isParatext "false" @default.
- W4311117974 isRetracted "false" @default.
- W4311117974 workType "article" @default.