Quantum-enhanced intrusion detection using multi-qubit entanglement
Conventional intrusion detection systems lack the ability to detect advanced cyberattacks because they rely on traditional computational models and linear correlations of features. This work proposes a simulation-based Quantum Intrusion Detection framework using the principles of quantum computing, with the main focus on multi-qubit entanglement and superposition to improve the detection of network anomalies. The multi-layered quantum circuit has three unique quantum state configurations: extended Bell states for 4-qubit pairwise correlations, Greenberger-Horne-Zeilinger (GHZ) states for 6-qubit multipartite entanglement and combined 8-qubit circuits with Quantum Fourier Transform for temporal analysis. The pipeline first converts classical network traffic features to quantum-compatible representations using statistical aggregation and correlation matrix construction. Cybersecurity datasets are handled by translating network features and grouping them into protocol type, packet size, interarrival time, source IP entropy, destination port, TCP flag, payload entropy, and flow duration using statistical aggregation. Quantum entanglement breaking is the main indicator of network anomaly, where malicious traffic patterns disturb the learned quantum correlations in baseline training done using normal traffic. The anomaly detection utilizes multi-level quantum analysis that integrates probability divergence measurement by Hellinger distance, entanglement deviation monitoring via concurrence computation and correlation breaking detection between quantum state pairs. The system uses weighted decision fusion with increased weight on 8-qubit unified circuit results. The system is evaluated through simulation using the NSL-KDD dataset, achieving an overall detection accuracy of 85%. This work serves as a baseline for investigating how quantum computing can be used in the field of cybersecurity in future quantum-enabled environments.