Loading...
Flexible deep reinforcement learning for beamforming optimization in 5g millimeter wave DF relay systems
Roensch, Walter
Roensch, Walter
Citations
Altmetric:
Files
Loading...
t25042_Roensch.pdf
Adobe PDF, 1.07 MB
Authors
Other Names
Location
Time Period
Advisors
Original Date
Digitization Date
Issue Date
2025-12-01
Type
Thesis
Genre
Keywords
Subjects (LCSH)
Electronic dissertations
Citation
Abstract
In this paper, a flexible deep reinforcement learning (DRL) approach to beamforming in a millimeter wave (mmWave) multiple-input multiple-output (MIMO) multi-cell relay network is proposed. This work investigate conditions in which users are dynamically allocated between cells with decode-and-forward (DF) relays, where users’ data are decoded and retransmitted. While near-optimal methods of beamforming may exist, they are computationally expensive. In this work, an intelligent beamforming paradigm using a multi-agent DRL and transfer learning is proposed. No work so far has provided a cooperative beamforming scheme for relays and source nodes (SNs) in the same RL scheme. Additionally, DRL beamforming methods have not been considered for inter-cell handovers in the literature. This work provides a novel method with RL using transferring learning to relieve performance drop due to retraining after handover. Simulation results verify that the proposed multi-cell DRL scheme outperforms traditional schemes by a wide margin.
Table of Contents
Description
Thesis (M.S.)-- Wichita State University, College of Engineering, Dept. of Electrical and Computer Engineering
Publisher
Wichita State University
Journal
Book Title
Series
Digital Collection
Finding Aid URL
Use and Reproduction
© Copyright 2025 by Walter Louis Roensch Jr.
All Rights Reserved