EUVIMEDEUVIMED entdecken
Open Access · CC BY 4.0

Revisiting Network Intrusion Detection: Benchmark Performance Analysis and Emerging Challenges

Abeer Rashad Mirdad

Journal of Intelligent Decision Making and Information Science · 09.09.2026 · 3 (18s) · 1622-1636

DOI: 10.17762/jidmis.v3.5885 ↗

Kurz gefasst

Fachartikel in der Zeitschrift Journal of Intelligent Decision Making and Information Science, online erschienen am 09.09.2026. Verfasst von Abeer Rashad Mirdad. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0).

AbstractBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Detection through intrusion detection systems (IDS) still dominates as a defense to mitigate network risks, whereas the development of IoT and industrial networks forces the detection to move from the traditional rule-based matching to learned normal behavior patterns. This review presents the existing literature divided into signature, anomaly-based (ML, DL, graph and federated models), and hybrid detection categories. Selected IDS are evaluated on the same benchmark datasets (NSL-KDD, UNSW-NB15, CICIDS2017, BoT-IoT) by their performance measures that are taken from original papers; therefore, no additional experiments are performed and the discussion is limited to the evaluation of what can or cannot be inferred from the numbers presented in publications. Three major points are observed throughout the analysis. First, accuracy exceeding 99% in many cases coexists with much lower values of multi-class or cross-representation performances on the same data. Therefore, it is unreasonable to use accuracy alone to evaluate models. Second, benchmark quality (duplicated instances, label errors, outdated data) influences comparability more than model architectures. Finally, the most interesting hybrids combine class-balancing generation, ensemble or attention-based classification, explainability and trust/federated layers. Finally, some future directions are formulated.

Zitieren

APA

Mirdad, A. R. (2026). Revisiting Network Intrusion Detection: Benchmark Performance Analysis and Emerging Challenges. Journal of Intelligent Decision Making and Information Science, 3(18s), 1622-1636. https://doi.org/10.17762/jidmis.v3.5885

Vancouver

Mirdad AR. Revisiting Network Intrusion Detection: Benchmark Performance Analysis and Emerging Challenges. Journal of Intelligent Decision Making and Information Science. 2026;3(18s):1622-1636. doi:10.17762/jidmis.v3.5885

BibTeX
@article{Mirdad2026,
  author = {Mirdad, Abeer Rashad},
  title = {Revisiting Network Intrusion Detection: Benchmark Performance Analysis and Emerging Challenges},
  journal = {Journal of Intelligent Decision Making and Information Science},
  year = {2026},
  volume = {3},
  number = {18s},
  pages = {1622--1636},
  doi = {10.17762/jidmis.v3.5885}
}
Kostenlos starten →