EUVIMEDEUVIMED entdecken
Crossref

An Intelligent Lightweight Hierarchical Intrusion Detection System Using Ensemble Decision Tree Techniques for WSNs

Shyam Sundar Selvam, Balaji Coimbatore GaneshSankar, Bhuvaneshwaran Raghuvel Subramaniam

Concurrency and Computation: Practice and Experience · 2026-07 · 38 (13)

DOI: 10.1002/cpe.70845 ↗

Kurz gefasst

Fachartikel in der Zeitschrift Concurrency and Computation: Practice and Experience, online erschienen am 09.07.2026. Verfasst von Shyam Sundar Selvam und 2 weiteren Autor:innen. Die Arbeit nennt 24 Literaturangaben.

AbstractBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

From environmental monitoring to industrial automation, wireless sensor networks (WSNs) form the backbone of numerous critical applications. Their inherent characteristics, like limited computational power, constrained memory, and finite energy reserves, render them exceptionally vulnerable to an increasingly sophisticated cyber‐attack. The existing intrusion detection systems (IDS) rely on computationally intensive machine learning (ML) or deep learning (DL) models that are not suited for the resource‐limited reality of WSN nodes. An intelligent framework specifically engineered for the operational constraints of WSNs, known as the L ightweight H ierarchical I ntrusion D etection S ystem ( L‐HIDS ), is introduced in this work. A two‐stage ensemble classifier was implemented in L‐HIDS, which collectively combines the benefits of the ID3 and C4.5 decision tree algorithms. The first stage utilizes an economical ID3 classifier for rapid, coarse‐grained detection, efficiently classifying between benign and malicious network traffic. Upon detecting a potential threat, the second stage activates an elaborate C4.5 classifier to perform fine‐grained classification, identifying the specific subtype of the attack. This hierarchical approach optimizes the trade‐off between detection accuracy and resource consumption. Extensive evaluations were conducted on two large‐scale datasets, namely CIC‐IDS2017 and BoT‐IoT, to authenticate L‐HIDS against modern threats. The L‐HIDS is benchmarked against standard ML models like Random Forest, XGBoost, and LightGBM, as well as a DL model (CNN‐LSTM). The observed results are convincing. L‐HIDS demonstrates highly competitive detection accuracy, precision, and F1‐scores. A significant reduction in energy consumption, up to 45% less than complex DL models, and lower packet delivery latency were achieved. The results confirm that L‐HIDS is a practical and highly effective solution for deploying robust, real‐time intrusion detection in practical WSN environments. It also addresses the critical gap between high‐level security and on‐device operational feasibility.

Zitieren

APA

Selvam, S. S., GaneshSankar, B. C., & Subramaniam, B. R. (2026). An Intelligent Lightweight Hierarchical Intrusion Detection System Using Ensemble Decision Tree Techniques for WSNs. Concurrency and Computation: Practice and Experience, 38(13). https://doi.org/10.1002/cpe.70845

Vancouver

Selvam SS, GaneshSankar BC, Subramaniam BR. An Intelligent Lightweight Hierarchical Intrusion Detection System Using Ensemble Decision Tree Techniques for WSNs. Concurrency and Computation: Practice and Experience. 2026;38(13). doi:10.1002/cpe.70845

BibTeX
@article{Selvam2026,
  author = {Selvam, Shyam Sundar and GaneshSankar, Balaji Coimbatore and Subramaniam, Bhuvaneshwaran Raghuvel},
  title = {An Intelligent Lightweight Hierarchical Intrusion Detection System Using Ensemble Decision Tree Techniques for WSNs},
  journal = {Concurrency and Computation: Practice and Experience},
  year = {2026},
  volume = {38},
  number = {13},
  doi = {10.1002/cpe.70845}
}
Kostenlos starten →