DC Field | Value | Language |
---|---|---|
dc.contributor.advisor | Chong, Song | - |
dc.contributor.advisor | 정송 | - |
dc.contributor.author | Bae, Jeongmin | - |
dc.date.accessioned | 2021-05-12T19:31:55Z | - |
dc.date.available | 2021-05-12T19:31:55Z | - |
dc.date.issued | 2019 | - |
dc.identifier.uri | http://library.kaist.ac.kr/search/detail/view.do?bibCtrlNo=886666&flag=dissertation | en_US |
dc.identifier.uri | http://hdl.handle.net/10203/283744 | - |
dc.description | 학위논문(석사) - 한국과학기술원 : 전기및전자공학부, 2019.2,[iv, 21 p. :] | - |
dc.description.abstract | As network architecture becomes complex and the user requirement gets diverse, the role of ecientnetwork resource management becomes more important. However, existing network scheduling algo-rithms such as the max-weight algorithm suer from poor delay performance. In this paper, we presenta reinforcement learning-based network scheduling algorithm that achieves both optimal throughput andlow delay. To this end, We rst formulate the network optimization problem as a dynamic programmingproblem. Then we introduce a new state-action value function called W-function and develop a rein-forcement learning algorithm called W-learning that guarantees little performance loss during a learningprocess. Finally, via simulation, we verify that our algorithm shows delay reduction of up to 40.8%compared to the max-weight algorithm over various scenarios. | - |
dc.language | eng | - |
dc.publisher | 한국과학기술원 | - |
dc.subject | Reinforcement learning▼anetwork utility optimization▼amax weight algorithm▼anetwork scheduling▼adynamic programming | - |
dc.subject | 강화 학습▼a네트워크 효용성 최대화 문제▼a최대 가중치 알고리즘▼a네트워크 스케줄링▼a동적 계획법 | - |
dc.title | Beyond max-weight scheduling | - |
dc.title.alternative | 강화 학습 기반 맥스-웨이트 스케줄링 개선 기법 | - |
dc.type | Thesis(Master) | - |
dc.identifier.CNRN | 325007 | - |
dc.description.department | 한국과학기술원 :전기및전자공학부, | - |
dc.contributor.alternativeauthor | 배정민 | - |
dc.title.subtitle | Reinforcement learning approach | - |
Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.