An investigation of feature reduction, transferability, and generalization in AWID datasets for secure Wi-Fi networks
2025

Detecting Cyberattacks in Wi-Fi Networks Using Machine Learning

publication 10 minutes Evidence: high

Author Information

Author(s): Khalid Nashmia, Hina Sadaf, Zaidi Khurram Shabih, Gaber Tarek, Speakman Lee, Noor Zainab

Primary Institution: Department of Computer Science, University of Engineering and Technology, Lahore, Pakistan

Hypothesis

Can a lightweight and cost-effective model effectively detect cyberattacks in Wi-Fi networks using feature reduction and machine learning techniques?

Conclusion

The study successfully developed a lightweight model that effectively detects various cyberattacks in Wi-Fi networks with high accuracy.

Supporting Evidence

  • The proposed model achieved 99.82% accuracy in detecting cyberattacks.
  • Recursive feature elimination was used to reduce the feature set from 16 to 8.
  • Decision trees and convolutional neural networks performed best in classification tasks.
  • The study highlighted the importance of feature transferability across different datasets.
  • Machine learning models were shown to be effective in real-time cyberattack detection.

Takeaway

This study shows how we can use smart computer programs to find bad guys trying to sneak into Wi-Fi networks and steal information.

Methodology

The study used recursive feature elimination with decision trees to select important features and tested various machine learning models for cyberattack detection.

Limitations

The study's findings may not generalize to wired networks, and the lack of proprietary datasets limits the scope of the research.

Digital Object Identifier (DOI)

10.1371/journal.pone.0306747

Want to read the original?

Access the complete publication on the publisher's website

View Original Publication