Predicting the Line-of-Sight Existence Using Radio Channel Properties and Relative Antenna Locations

Katsuyuki Haneda*, Rachana Desai

*Corresponding author for this work

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

100 Downloads (Pure)

Abstract

Existence of the line-of-sight (LoS) in a radio link is an important feature determining radio link and localization performance. This study explores relationship between multidimensional nature of radio channel properties such as delay and angular spreads and link pathloss with the LoS existence along with the relative antenna geometry of the two communicating devices in an attempt to predict the LoS existence. A nonlinear classification method of the decision tree is trained by extensive wideband double-directional channel sounding data in a microcellular environment. It was found that the BS-MS link distance and link pathloss were most influential features in predicting the LOS existence. While the use of less influential features of delay and angular spreads led to 75 % accuracy due to not being able to find the exact pattern across different MS routes. We thereby demonstrate the usefulness of incorporating geometrical knowledge of the link to predict the LoS state, despite not using a map of the cellular site, in addition to using channel parameters.

Original languageEnglish
Title of host publication17th European Conference on Antennas and Propagation, EuCAP 2023
PublisherIEEE
ISBN (Print)978-1-6654-7541-9
DOIs
Publication statusPublished - 2023
MoE publication typeA4 Conference publication
EventEuropean Conference on Antennas and Propagation - Florence, Italy, Florence, Italy
Duration: 26 Mar 202331 Mar 2023
Conference number: 17
https://www.eucap2023.org/

Conference

ConferenceEuropean Conference on Antennas and Propagation
Abbreviated titleEuCAP
Country/TerritoryItaly
CityFlorence
Period26/03/202331/03/2023
Internet address

Fingerprint

Dive into the research topics of 'Predicting the Line-of-Sight Existence Using Radio Channel Properties and Relative Antenna Locations'. Together they form a unique fingerprint.

Cite this