Fortifying the Connection: Cybersecurity Tactics for WSN-driven Smart Manufacturing in the Era of Industry 5.0

Abstract

Wireless Sensor Network (WSN)-based manufacturing facilities in the context of the Fourth Industrial Revolution (Industry 5.0) represent advanced Cyber-Physical Production Systems (CPPSs), wherein seamless networking of people, objects, and machines is achieved across the entire supply chain. A significant advantage of such digitization is the facilitation of personalized and agile manufacturing processes. However, this interconnectedness introduces a spectrum of novel threat vectors, enabling sophisticated Distributed Denial-of-Service (DDoS) attacks. One critical vulnerability lies in the Internet of Things (IoT) sensor nodes. These IoT devices, now extensively utilized for sensing, data acquisition, analysis, and communication within manufacturing environments, have concomitantly escalated the risk of cyber threats. To counteract these threats, advanced intrusion detection systems leveraging deep learning algorithms have emerged as scalable and intelligent solutions for safeguarding industrial IoT and WSN infrastructures. This paper introduces a hybrid Convolutional Neural Network-Long Short-Term Memory (CNN-LSTM) model tailored for cybersecurity attack detection in industrial IoT environments, specifically within WSN-based smart manufacturing contexts. The proposed CNN-LSTM model exhibits superior efficacy in identifying DDoS attacks within Industry 5.0 CPS environments, surpassing conventional Artificial Neural Network (ANN), Long Short-Term Memory (LSTM), and Gated Recurrent Unit (GRU) models in terms of accuracy, precision, recall, and F1-score. Utilizing real-world network traffic datasets, the developed deep learning-based network anomaly detection system enhances the capability to detect and mitigate cyber threats, thereby reinforcing the security and resilience of smart manufacturing systems. The practical benefits of this enhanced cyberattack detection system include improved operational reliability, reduced downtime, and the protection of critical assets in real-world smart manufacturing settings.