A hybrid SHA-256 and variational quantum classifier framework for simulation-based intrusion detection in SDN-IoT networks

Abstract

Abstract The convergence of Software-Defined Networking (SDN) with Internet of Things (IoT) introduces critical security vulnerabilities requiring adaptive detection mechanisms. This paper presents QEIDS, a quantum-enhanced intrusion detection system combining SHA-256 cryptographic fingerprinting with variational quantum classification (VQC) for efficient threat detection in SDN-IoT networks. The three-layer architecture employs SHA-256 hash matching for deterministic primary filtering (handling 60–70% of traffic at 4,000 packets/s), a Variational Quantum Classifier using EfficientSU2 ansatz with ZFeatureMap encoding on 9 qubits for adaptive threat classification of ambiguous traffic, and a decision-tree fallback for interpretable edge-case handling. We establish polynomial sample complexity bounds of O(n³ log n) for the VQC-based learning framework with strategic initialization, compared to higher polynomial or exponential worst-case bounds for partition function computation in classical energy-based models. Extensive evaluation across four benchmark datasets NSL-KDD (148,517 samples, network intrusions), PhiUSIIL (235,162 samples, phishing URLs), Android Permissions (29,932 samples, malware), and Email Spam (5,572 samples) demonstrates 97.34% mean detection accuracy with 35% reduction in computational footprint compared to classical deep learning baselines. The hybrid framework achieves 96.8 ms average detection latency under MATLAB 2024b classical simulation, suggesting potential compliance with the sub-100 ms SDN flow-processing threshold; however, hardware validation on actual NISQ devices remains as future work. Additionally, QEIDS attains 99.2% recall on critical attack types (DDoS, probe, and zero-day exploits), 84.27% average zero-shot accuracy on previously unseen attack patterns, and 4.51× faster training convergence compared to CNN-GRU baselines. All results are simulation-based and represent proof-of-concept validation; practical quantum advantage claims require future verification on real quantum hardware.