Deep-learning-based Estimation of Radio-quality Deterioration Causes for 5G Industrial Applications

It has become increasingly important for industry to promote productivity by utilizing 5G and industrial internet of things (IIoT). However, radio quality is difficult to assure due to deteriorations such as shadowing and fading. A means to automatically identify the root causes of radio-quality det...

Full description

Saved in:
Bibliographic Details
Main Authors: Nishikawa, Yoshiaki, Maruyama, Shohei, Takahashi, Eiji, Onishi, Takeo
Format: Conference Proceeding
Language:English
Subjects:
deep learning
fading
radio quality deterioration
root cause identification
shadowing
Online Access:Request full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:It has become increasingly important for industry to promote productivity by utilizing 5G and industrial internet of things (IIoT). However, radio quality is difficult to assure due to deteriorations such as shadowing and fading. A means to automatically identify the root causes of radio-quality deterioration is expected to enable prompt measures. This paper proposes a method to estimate causes of radio-quality deterioration by using a deep learning model and the reference signal received power (RSRP). The method has two key features: i) it uses only one of the easiest kinds of data to retrieve; ii) it only uses simulated data for training. Evaluation experiments were conducted on a private 5G network in an operating factory. The method achieved about 0.95 in F1 score. This indicates that our trained-by-simulation model can work in a real situation.
ISSN:2331-9860
DOI:10.1109/CCNC51644.2023.10060500