Efficient TIS Sensitivity Measurement with Machine Learning Approach and 5G Dataset

Yi Wei Chen, Min Je Tsai, Henry Horng Shing Lu, Kai Ten Feng, Ta Sung Lee, Jih Chuan Lan

Research output: Chapter in Book/Report/Conference proceedingConference contributionpeer-review

Abstract

Total isotropic sensitivity (TIS) measurement is strongly required by the industry, but the procedure takes a long time. We explore a machine learning (ML) approach to speed up TIS test procedure. The experiments are conducted using 5G devices and frequency bands. The results show that our methodology can improve measurement efficiency by 35% to 65%, while still maintain high accuracy within 1 dB deviation from standard procedure. Disconnection is a critical issue during TIS measurement, so we design a calibration mechanism to reduce the risk of disconnection. Our approach can be applied widely to different system configurations. It supports not only 5G but also previous generations on different frequency bands.

Original languageEnglish
Title of host publication2024 IEEE 21st Consumer Communications and Networking Conference, CCNC 2024
PublisherInstitute of Electrical and Electronics Engineers Inc.
Pages1032-1033
Number of pages2
ISBN (Electronic)9798350304572
DOIs
StatePublished - 2024
Event21st IEEE Consumer Communications and Networking Conference, CCNC 2024 - Las Vegas, United States
Duration: 6 Jan 20249 Jan 2024

Publication series

NameProceedings - IEEE Consumer Communications and Networking Conference, CCNC
ISSN (Print)2331-9860

Conference

Conference21st IEEE Consumer Communications and Networking Conference, CCNC 2024
Country/TerritoryUnited States
CityLas Vegas
Period6/01/249/01/24

Keywords

  • 5G
  • machine learning
  • sensitivity measurement
  • TIS

Fingerprint

Dive into the research topics of 'Efficient TIS Sensitivity Measurement with Machine Learning Approach and 5G Dataset'. Together they form a unique fingerprint.

Cite this