A Sustainable Quantum-Enhanced and Resource-Efficient Ensemble Model for Robust Man-in-the-Middle Attack Detection

Introduction/Objective: Man-in-the-Middle (MITM) attacks are a major threat to both the privacy and security of network communications because they operate in secrecy while taking advantage of weak encryption methods, inadequate security systems, and session routing. This research develops an intrusion detection system that can effectively detect MITM attacks through its advanced detection capabilities and ability to handle multiple datasets and machine learning system requirements, which require less computational power. Methods: The research presents a Quantum-Enhanced and Resource-Efficient Ensemble Framework, which combines Quantum Genetic Algorithm-based Feature Selection (QGA-FS) with a hybrid quantum-classical Variational Quantum Classifier (VQC). QGA-FS implements qubit-encoded feature representation together with rotation-gate evolution to identify essential features that require low computational resources. The optimized feature set is then classified using a VQC to capture complex nonlinear relationships. The quantum components combine with Random Forest, AdaBoost, and XGBoost to create a system that improves both system stability and user understanding. The UNSW-NB15 and CICIDS2017 datasets serve as testing grounds for the framework, which will be assessed through cross-dataset validation. Results: The experimental results show that the framework achieves a 98.4% mean detection accuracy with an F1-score of 0.982 across all tested datasets. The model achieves energy savings of 31% when compared to traditional ensemble-based intrusion detection systems, which demonstrates its ability to operate efficiently. Discussion: The findings demonstrate that using quantum-inspired feature optimization together with hybrid quantum-classical classification and ensemble learning results in better detection accuracy while using minimal computational and energy resources. The proposed framework demonstrates its ability to operate effectively across different network environments according to cross-dataset validation results. Conclusion: The Quantum-Enhanced and Resource-Efficient Ensemble Framework presents an effective and sustainable solution for MITM attack detection according to the results of this study. The framework shows how quantum-assisted learning can create cyber-resilient intrusion detection systems, which also prepare for future quantum threats, because it combines high detection accuracy with flexibility and low energy consumption.