Skip to content
DSML.KZ
Publications

Code · Code Ocean · August 11, 2026

An Adaptive Zero-Trust Blockchain Framework for Low-Latency, Privacy-Preserving, and Explainable Intrusion Detection in Industrial IoT

Ali Akhmetkaliyev, Mohammed Alaa Ala’anzy, Yassine Himeur

Mohammed Alaa Fadhil Al-hadeethi
Open via DOI

Abstract

This project introduces an end-to-end Adaptive Zero-Trust Blockchain Framework designed for low-latency, privacy-preserving, and explainable intrusion detection in Industrial IoT (IIoT) environments. Built to operate under OT timing and resource constraints, the framework integrates sequential machine learning detection with continuous trust verification, tamper-evident logging, and operator triage visualization. Core Architecture & Pipeline ComponentsLightweight Detection Engine: Utilizes a feature selection method paired with a CNN–BiLSTM architecture. By retaining Mutual Information (MI) feature filtering while removing computationally heavy Dual-Attention mechanisms, the pipeline achieves a 39% latency reduction (0.11 s for 500 samples) while maintaining 98% overall accuracy and 100% recall on critical attack classes.Laplace Differential Privacy Module: Incorporates Laplace Differential Privacy ($\varepsilon = 50.0$) at training time to regularize noise and protect against privacy leakage while maintaining model utility ($\le 0.5$ percentage point accuracy drop).Zero Trust Manager & Hash-Chained Ledger: Replaces heavy consensus smart contracts with a lightweight SHA-256 hash-chained ledger operating at $\mathcal{O}(1)$ append complexity. Device trust scores continuously adjust based on model confidence ($T = \text{clip}(T + (c - 0.5) \times 0.5)$).Digital Twin & Explainability Triage: Translates real-time device trust scores into dynamic operational states (Healthy, Degraded, Quarantined) on a Digital Twin dashboard. Explainability is provided using a KernelExplainer SHAP pairwise interaction matrix for immediate operator triage.

Blockchain Technology Applications and SecurityNetwork Security and Intrusion DetectionInternet of Things and AIzero-trust securityblockchainIndustrial Internet of Things

More by Mohammed Alaa Fadhil Al-hadeethi