Title

Revealing Insights for Small Cell Networks via Optimization and Deep Reinforcement Learning

Alternative Title

스몰 셀 네트워크에서 효용 함수 최대화를 위한 최적화 및 심층 강화학습 기반 네트워크 자원 관리 알고리즘 개발

DGIST Authors

Pildo Yoon ; Jeongho Kwak ; Daewon Seo

Advisor

곽정호

Co-Advisor(s)

Daewon Seo

Issued Date

2024

Awarded Date

2024-02-01

Citation

Pildo Yoon. (2024). Revealing Insights for Small Cell Networks via Optimization and Deep Reinforcement Learning. doi: 10.22677/THESIS.200000724592

Type

Thesis

Description

스몰 셀 네트워크;자원 관리;간섭 관리;Lyapunov optimization 최적화;강화학습

Table Of Contents

1. Introduction
1.1 Expected Future Network and Challenges
1.2 Motivation and Contributions
2. ULTIMA: Ultimate Balance of Centralized and Distributed Benefits for Interference Management in 5G Cellular Networks
2.1 Introduction
2.2. Related Works
2.3 EdgeSON architecture
2.4 System Model
2.4.1 Resource and Allocation Model
2.4.2 Link Model
2.5 Multi-tier IMPowerShare Algorithm
2.5.1 Problem Formulation
2.5.2 Algorithm Development
2.5.3 Algorithm Description
2.6. Performance Analysis
2.6.1 Simulation Results
2.7 Discussion
2.8 Conclusion
2.9 Appendix
2.9.1 Proof of Lemma 1.
3. Judgement-based Deep Q-Learning Framework for Interference Management in Small Cell Networks
3.1 Introduction
3.2 System Model and Problem Formulation
3.3 Multi-Agent DQN Approach with Novel Architecture and Reward Design
3.3.1 Deep Q-Network Architecture Design
3.3.2 State, Action and Reward Design
3.4 Simulation Results and Analysis
3.5 Conclusion

Reference

URI

http://hdl.handle.net/20.500.11750/48108
http://dgist.dcollection.net/common/orgView/200000724592

DOI

10.22677/THESIS.200000724592

Degree

Master

Department

Department of Electrical Engineering and Computer Science

Publisher

DGIST

Show Full Item Record