Related Experiment Video
Updated: May 29, 2025
08:05
Design and Analysis for Fall Detection System Simplification
Published on: April 6, 2020
10.6K
A hybrid machine learning model for intrusion detection in wireless sensor networks leveraging data balancing and
Md Alamin Talukder1, Majdi Khalid2, Nasrin Sultana3
1Department of Computer Science and Engineering, International University of Business Agriculture and Technology, Dhaka, Bangladesh. alamin.cse@iubat.edu.
Scientific Reports
|February 7, 2025
Summary
This study introduces a hybrid machine learning model for robust intrusion detection in Wireless Sensor Networks (WSNs) and the Internet of Things (IoT). The model significantly enhances accuracy and efficiency, outperforming existing methods.
Area of Science:
- Cybersecurity
- Machine Learning
- Network Security
Background:
- Wireless Sensor Networks (WSNs) and Internet of Things (IoT) environments face significant security challenges from various threats.
- Effective intrusion detection systems (IDS) are crucial for safeguarding these interconnected systems.
- Traditional methods often struggle with data imbalance and high dimensionality, impacting detection performance.
Purpose of the Study:
- To develop a novel hybrid machine learning model for enhanced intrusion detection in WSNs and IoT.
- To address the challenges of class imbalance and high dimensionality in network traffic data.
- To improve the accuracy, efficiency, and scalability of intrusion detection systems.
Main Methods:
- Integration of KMeans-SMOTE (KMS) for effective data balancing.
- Application of Principal Component Analysis (PCA) for dimensionality reduction.
- Utilizing classifiers like Decision Tree, Random Forest Classifier (RFC), and XGBoost (XGBC).
Main Results:
- The hybrid (KMS + PCA + RFC) model achieved exceptional accuracy (99.94% on WSN-DS, 99.97% on TON-IoT).
- The model demonstrated superior performance compared to traditional SMOTE TomekLink and Generative Adversarial Network (GAN) balancing techniques.
- Complexity analysis confirmed reduced training and prediction times, indicating suitability for real-time applications.
Conclusions:
- The proposed hybrid ML model offers a scalable and robust solution for intrusion detection in WSNs and IoT.
- This approach effectively mitigates class imbalance and high-dimensionality issues.
- The model's high accuracy and efficiency make it a promising advancement for network security.
