USOD: A Unified Security Operations Dashboard with Hybrid AI Intrusion Detection and Immutable Blockchain Audit Trails
Authors: Ghulam Mohayudin, Ali Sami, Zuhaib Iqbal, Iftikhar Rasheed
Abstract
Network intrusion detection systems often suffer from a critical gap between threat identification and immutable audit logging. We introduce USOD (Unified Security Operations Dashboard), a microservices-based architecture that bridges this gap by coupling a hybrid XGBoost/Isolation Forest inference engine with a Hyperledger Fabric blockchain. Unlike traditional SIEMs that rely on centralized, mutable databases, USOD enforces cryptographic non-repudiation for high-severity alerts. Our evaluation on the CICIDS2017 dataset demonstrates that XGBoost achieves 99.88% accuracy, while the Raft-consensus blockchain commits threat logs with a mean latency of 47ms. This dual-track approach combines probabilistic detection with deterministic auditing, providing a verified, tamper-proof security posture accessible via synchronized Web, Mobile, and Desktop clients.
