Explainable AI Techniques for Network Traffic Anomaly Detection and Mitigation: A Systematic Review.
IOSR Journal of Computer Engineering · 2026-05 · 28 (5) · 01-11
DOI: 10.9790/0661-2805020111 ↗Kurz gefasst
Systematische Übersicht in der Zeitschrift IOSR Journal of Computer Engineering, 2026-05. Verfasst von Fumlack Kingsley George und 5 weiteren Autor:innen.
Abstract (Originalsprache)
The modern networks have become more complex making it difficult to detect anomalies in real time and reduce congestion. Conventional machine learning models, however precise, are black boxes, which do not provide much information about the way they make decisions. To solve this, Explainable Artificial Intelligence (XAI) offers transparency, interpretability and accountability in network monitoring. This paper locate, assess and contrast XAI methods used in network anomaly detection with a particular focus on model transparency and point out the research gaps. Research articles published in 2020-2025 in IEEE Xplore, Scopus, SpringerLink, and ScienceDirect were systematic review articles that are peer-reviewed and analyzed using the PRISMA framework. The results demonstrate that SHAP, LIME, Grad-CAM, attention mechanisms and rule-based explanations have been the most used, each with its strengths and trade-offs of computation. Model-agnostic approaches have greater interpretability, but are computationally expensive, whereas model-specific approaches are architecture-specific, and thus can explain faster. Real-time explainable frameworks and standardized interpretability metrics are some of the research priorities identified in the review. In general, explainability is also placed as a fundamental pillar of trustful AI, which positively influences transparency, user trust and responsibility of intelligent network management systems
Zitieren
APA
George, F. K., Bethel, N. U., Gago, K. B., AGBOGUN, J. B., Baba, U. M., & Mohammed, M. (2026). Explainable AI Techniques for Network Traffic Anomaly Detection and Mitigation: A Systematic Review. IOSR Journal of Computer Engineering, 28(5), 01-11. https://doi.org/10.9790/0661-2805020111
Vancouver
George FK, Bethel NU, Gago KB, AGBOGUN JB, Baba UM, Mohammed M. Explainable AI Techniques for Network Traffic Anomaly Detection and Mitigation: A Systematic Review. IOSR Journal of Computer Engineering. 2026;28(5):01-11. doi:10.9790/0661-2805020111
BibTeX
@article{George2026,
author = {George, Fumlack Kingsley and Bethel, Nnamani Ugochukwu and Gago, Kindness Bulus and AGBOGUN, Joshua Babatunde and Baba, Uloko Matthew and Mohammed, Murtala},
title = {Explainable AI Techniques for Network Traffic Anomaly Detection and Mitigation: A Systematic Review},
journal = {IOSR Journal of Computer Engineering},
year = {2026},
volume = {28},
number = {5},
pages = {01--11},
doi = {10.9790/0661-2805020111}
}
EUVIMED