Vision-Position Multi-Modal Beam Prediction Using Real Millimeter Wave Datasets

Gouranga Charan, Tawfik Osman, Andrew Hredzak, Ngwe Thawdar, Ahmed Alkhateeb

Research output: Chapter in Book/Report/Conference proceedingConference contribution

23 Scopus citations

Abstract

Enabling highly-mobile millimeter wave (mmWave) and terahertz (THz) wireless communication applications requires overcoming the critical challenges associated with the large antenna arrays deployed at these systems. In particular, adjusting the narrow beams of these antenna arrays typically incurs high beam training overhead that scales with the number of antennas. To address these challenges, this paper proposes a multi-modal machine learning based approach that leverages positional and visual (camera) data collected from the wireless communication environment for fast beam prediction. The developed framework has been tested on a real-world vehicular dataset comprising practical GPS, camera, and mmWave beam training data. The results show the proposed approach achieves more than 75% top-1 beam prediction accuracy and close to 100% top-3≈beam prediction accuracy in realistic communication scenarios.

Original languageEnglish (US)
Title of host publication2022 IEEE Wireless Communications and Networking Conference, WCNC 2022
PublisherInstitute of Electrical and Electronics Engineers Inc.
Pages2727-2731
Number of pages5
ISBN (Electronic)9781665442664
DOIs
StatePublished - 2022
Event2022 IEEE Wireless Communications and Networking Conference, WCNC 2022 - Austin, United States
Duration: Apr 10 2022Apr 13 2022

Publication series

NameIEEE Wireless Communications and Networking Conference, WCNC
Volume2022-April
ISSN (Print)1525-3511

Conference

Conference2022 IEEE Wireless Communications and Networking Conference, WCNC 2022
Country/TerritoryUnited States
CityAustin
Period4/10/224/13/22

ASJC Scopus subject areas

  • General Engineering

Fingerprint

Dive into the research topics of 'Vision-Position Multi-Modal Beam Prediction Using Real Millimeter Wave Datasets'. Together they form a unique fingerprint.

Cite this