Industrial Wi-Fi Traffic Forecasting Using Diverse Time-Series Models

Citations

SCOPUS

1

초록

Smart factories cannot be detached from wireless networks, which have become essential elements for real-time communication and control of various industrial devices. Accurate traffic prediction is indispensable for managing network resources efficiently, enabling proactive resource allocation, congestion avoidance, and load balancing. In this paper, we conduct an evaluation based on actual measurement data obtained from Wi-Fi Access Points (APs) and Automated Guided Vehicles (AGVs) in a real automobile manufacturing factory. We predict the AP’s traffic volume using various deep learning models, including Multi-Layer Perceptron (MLP), Long Short-Term Memory (LSTM), Transformer, and Graph Neural Network (GNN) architectures, for future traffic forecasting. Our results show that the independent models, which are trained for each AP, effectively capture individual traffic patterns.

키워드

Deep Learning; Time Series Forecasting; Traffic Prediction; Wi-Fi; Wireless Network
제목
Industrial Wi-Fi Traffic Forecasting Using Diverse Time-Series Models
저자
Choi, Minjun; Shin, Jae Hong; Paek, Jeongyeup
DOI
10.1109/ICTC66702.2025.11388511
발행일
2025
유형
Conference Paper
저널명
International Conference on ICT Convergence
페이지
671 ~ 674