Copyright © 2026 Authors retain the copyright of this article. This article is an open access article distributed under the Creative Commons Attribution License which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

@article{205202,
        author = {Anirudha Anil Gaikwad and Atit Anil Gaikwad},
        title = {Quantum-Enhanced Hybrid Anomaly Detection for Proactive Cybersecurity in Distributed Networks},
        journal = {International Journal of Innovative Research in Technology},
        year = {2026},
        volume = {13},
        number = {1},
        pages = {5755-5766},
        issn = {2349-6002},
        url = {https://ijirt.org/article?manuscript=205202},
        abstract = {The distributed network infrastructure of today is facing a continuously growing attack surface, and the traditional intrusion detection systems are not able to cope up with the high false alarm rates and slow response against polymorphic attacks. In this work, the authors have proposed QHybrid-AD, a quantum-classical anomaly detection framework in which a Variational Autoencoder (VAE) front-end is combined with a Quantum Support Vector Machine (QSVM) classifier that operates in a Hilbert-space feature representation generated through the ZZ-entangling feature map. In the first step, the VAE compresses high-dimensional and noisy traffic data into a compact latent vector. Thereafter, the QSVM uses parameterised quantum circuits on a simulated 8-qubit backend for separating normal flows from attacks. The proposed framework has been evaluated on NSL-KDD and CIC-IDS-2017 datasets under realistic noise conditions calibrated to IBM Eagle r3 specifications. QHybrid-AD achieves an Accuracy of 97.84% and F1-score of 96.91% on NSL-KDD binary classification, together with a 27.3% reduction in false positives when compared against a tuned radial-basis-function SVM. On CIC-IDS-2017, the framework gives 96.52% Accuracy with a 23.8% reduction in false alarms. Scalability tests on a simulated 12-node distributed topology show sub-linear growth in the detection latency, which confirms the practical viability for edge-proximate deployment. Ablation studies have been performed for separating the contributions of the quantum kernel, the VAE bottleneck width, and the zero-noise extrapolation. The results show that the ZZ-feature map is capable of capturing the inter-feature correlations that are not accessible to classical polynomial and Gaussian kernels. Further, the quantum noise mitigation recovers up to 4.1 percentage points of Accuracy which is otherwise lost under depolarising error channels.},
        keywords = {Quantum machine learning, anomaly detection, intrusion detection system, quantum support vector machine, variational autoencoder, ZZFeatureMap, distributed networks, NISQ.},
        month = {June},
        }