Industrial Wireless Cyberphysical Systems Performance Using Deep Learning

Published

September 28, 2023

Author(s)

Mohamed Hany, Rick Candell, Karl Montgomery

Abstract

Industrial wireless communications networks have a major role in the future industrial cyber-physical systems (CPSs) to have higher flexibility and massive machine connectivity. However, the impact of the industrial wireless on the reliability and latency needs to be evaluated under various scenarios in order to meet the industrial CPS requirements. In order to measure the impact of deploying industrial wireless on cyber-physical system performance, a deep learning framework based on the generative adversarial network (GAN) is introduced. The GAN is used to model the performance of the system and identify the impact of various wireless impairments on the system performance. The GAN model can include features from both the wireless network and operational performance spaces. The GAN loss function is deployed as a performance metric that can be used to both model the average performance of a specific wireless scenario and identify specific wireless events that degrade the instantaneous industrial wireless performance during the operation of an industrial wireless system. The proposed GAN methodology is validated using a dual-robot collaborative lift use case in which IEEE 802.11ac is employed.

Proceedings Title

Manufacturing Science and Engineering Conference MSEC 2023

Conference Dates

June 12-16, 2023

Conference Location

New Brunswick, NJ, US

Pub Type

Conferences

Download Paper

https://doi.org/10.1115/MSEC2023-100969

Local Download

Keywords

industrial wireless, deep learning, generative adversarial networks, performance evaluation

Citation

Hany, M. , Candell, R. and Montgomery, K. (2023), Industrial Wireless Cyberphysical Systems Performance Using Deep Learning, Manufacturing Science and Engineering Conference MSEC 2023 , New Brunswick, NJ, US, [online], https://doi.org/10.1115/MSEC2023-100969, https://tsapps.nist.gov/publication/get_pdf.cfm?pub_id=935814 (Accessed October 13, 2025)