ML-Enabled Eavesdropper Detection in Beyond 5G IIoT Networks

TL;DR

Eavesdropping threats in beyond-5G IIoT networks are addressed with AI driven physical layer security using CSI based eavesdropper detection. The authors simulate a dense industrial B5G system and evaluate RF, DCNN, DCNN 2, and LSTM models trained on CSI images (and optional position/power data) to classify UEs as legitimate or eavesdroppers, reporting high detection accuracy and analyzing secrecy rates. RF and DCNN 2 emerge as top performers, with RF often yielding higher secrecy rate under test density, and DCNN 2 benefiting from data fusion in improved detection. The work demonstrates the potential of AI enhanced PLS for secure B5G IIoT networks and provides a framework for evaluating CSI based authentication in industrial settings.

Abstract

Advanced fifth generation (5G) and beyond (B5G) communication networks have revolutionized wireless technologies, supporting ultra-high data rates, low latency, and massive connectivity. However, they also introduce vulnerabilities, particularly in decentralized Industrial Internet of Things (IIoT) environments. Traditional cryptographic methods struggle with scalability and complexity, leading researchers to explore Artificial Intelligence (AI)-driven physical layer techniques for secure communications. In this context, this paper focuses on the utilization of Machine and Deep Learning (ML/DL) techniques to tackle with the common problem of eavesdropping detection. To this end, a simulated industrial B5G heterogeneous wireless network is used to evaluate the performance of various ML/DL models, including Random Forests (RF), Deep Convolutional Neural Networks (DCNN), and Long Short-Term Memory (LSTM) networks. These models classify users as either legitimate or malicious ones based on channel state information (CSI), position data, and transmission power. According to the presented numerical results, DCNN and RF models achieve a detection accuracy approaching 100\% in identifying eavesdroppers with zero false alarms. In general, this work underlines the great potential of combining AI and Physical Layer Security (PLS) for next-generation wireless networks in order to address evolving security threats.

Figures (5)

• Figure 1: IIoT network topology
• Figure 2: Proposed DCNN's Structure
• Figure 3: Proposed LSTM's Structure
• Figure 4: Training Time Comparison of Different Models
• Figure 5: Average Secrecy Rate of system for DCNN 2 and RF