Adaptive Machine Learning Algorithms for mmWave Communications in Beyond 5G and 6G Systems

2023 Fiscal Year Final Research Report

• Project/Area Number: 21K14162
• Research Category: Grant-in-Aid for Early-Career Scientists
• Allocation Type: Multi-year Fund
• Review Section: Basic Section 21020:Communication and network engineering-related
• Research Institution: Institute of Physical and Chemical Research
• Principal Investigator: Hashima Sherief 国立研究開発法人理化学研究所, 革新知能統合研究センター, 研究員 (00865462)
• Project Period (FY): 2021-04-01 – 2024-03-31
• Keywords: D2D Communications / millimeter Wave (mmWave) / Radio frequency (RF) / Multi-armed bandit (MAB) / Metaverse / Thompson sampling (TS) / UAV / V2X communications

Outline of Final Research Achievements

In this project, we modeled mathematical formulations for different wireless communication scenarios using online learning algorithms including Neighborhood discovery and selection in millimeter wave D2D communications, Gateway UAV selection in disaster area scenario, RIS aided mmWave Communications, UAV mounted RIS, Multiband wireless networks, V2X over the air updates, and V2X metaverse content updates. We leverged various types of multi armed bandits to these scenarios and modified it to be energy aware. The Utilized Bandit solutions showed near optimal performance and fast convergence rate in various settings due to proper selection and their unique learning policy. As a proof of concept we conducted experimental setup to memic the scenario of over the air updates using smart road side units.

Free Research Field

Wireless Communications

Academic Significance and Societal Importance of the Research Achievements

Thr project contributes to fullfilling the requirments of advanced communication technology(6G) with effective energy aware performance. The project also has social impact in disaster relief as it ensured reliable communications in challenging environments. Also,it contributes to autonomus vehicles