CONQUEST: CONtext-Aware QUantum-Inspired Intelligence for Espying AI-aided Cyberworm Security Threats in Military Networks

초록

Military networks face AI-aided cyberworms capable of adaptive, stealthy intrusions. We propose CONQUEST, a context-aware, quantum-inspired framework integrating Quantum-Inspired Particle Swarm Optimization (QPSO) for high-dimensional feature selection, LSTM-based Variational Autoencoder (LSTM-VAE) for anomaly detection, and operational context from the ACI-IoT dataset. QPSO reduces inference latency while preserving accuracy, and contextual analysis enhances detection precision. Evaluated on ACI-IoT, CONQUEST achieved a 97.5% detection rate with 0.8% false positives, surpassing signature, statistical, and conventional AI methods. These results demonstrate CONQUEST's potential for real-time, scalable defense in mission-critical military communications against advanced AI-driven cyber threats.