Volume 17, Number 3

Wireless Sensor Networks, Intrusion Detection, Stacking Ensemble Learning, Optuna, Feature Selection, XGB, CatBoost.

  Authors

Dilip Dalgade, NileshPatil, ManujJoshi and Dilendra Hiran, Pacific Academy of Higher Education and Research, India

  Abstract

Wireless Sensor Networks (WSNs) are key for ubiquitous computing. Despite advantages, they face security challenges due to decentralized nature and threats. Intrusion detection helps protect WSNs from security threats. This study proposes an Optuna-implemented stacking technique (OXCRF) the method combines SHapley Additive exPlanations, CatBoost, Mutual Information, and cross-validated Recursive Feature Elimination with Random Forest for feature selection, while SMOTE handles data imbalance. The stacking ensemble, XGBoost, CatBoost and Random Forest are used as the base learners, with hyperparameters being optimized using Optuna. Experiments on the NSL-KDD and UNSW-NB15 datasets show that OXCRF achieves higher accuracy (99.60% for binary and 99.53% for multiclass on NSL-KDD; 98.62% for binary and 83.67% for multiclass on UNSW-NB15) and lower misclassification rates (0.0040 and 0.0047 on NSLKDD; 0.0138 and 0.1633 on UNSW-NB15) compared to baseline models. Running an ablation study showed that OXCRF components worked as expected for multiclass intrusion detection in WSNs with overlapping classes and imbalanced data. The framework is efficient through feature selection, balanced data distribution and improved ensemble learning.

  Keywords

Wireless Sensor Networks, Intrusion Detection, Stacking Ensemble Learning, Optuna, Feature Selection, XGB, CatBoost.