Explainable Blockchain-Enabled Intrusion Detection Framework for Secure and Trustworthy 5G-IIoT Networks

Abstract

The integration of 5G networks and the Industrial Internet of Things (IIoT) enables real-time industrial automation but also expands the cybersecurity attack surface. Although previous studies have proposed AI and blockchain-based security frameworks, intrusion detection in 5G-IIoT remains limited by black-box AI models, low interpretability, and blockchain mechanisms that mainly support logging rather than attack detection. This study proposes an Explainable Blockchain-Enabled Intrusion Detection System (XB-IDS) for secure 5G-IIoT networks. The framework integrates deep learning-based intrusion detection, SHAP-based explainability, and blockchain-enabled security logging with smart contracts. A hybrid CNN-LSTM model is used to detect spatial and temporal attack patterns, while SHAP provides interpretable explanations for security analysts. Public IIoT cybersecurity datasets such as TON_IoT, Edge-IIoTset, and CICIoT2023 are used for evaluation. The proposed framework is assessed using accuracy, precision, recall, F1-score, false positive rate, detection latency, throughput, and explainability analysis. The proposed XB-IDS aims to improve detection performance, transparency, and trustworthiness in 5G-IIoT security operations. This study contributes an experimentally evaluable framework that extends prior AI-blockchain security research toward explainable and accountable intrusion detection.