Firefly algorithm based WSN-IoT security enhancement with machine learning for intrusion detection

Area of Science:

• Cybersecurity
• Network Security
• Machine Learning Applications

Background:

• Wireless Sensor Networks (WSN) and the Internet of Things (IoT) are increasingly integrated for enhanced data analysis and automation.
• Securing these interconnected WSN and IoT systems is critical for reliability and safety.
• Existing security measures require advancement to address the complexities of integrated WSN-IoT environments.

Purpose of the Study:

• To develop and evaluate a novel security technique for WSN-IoT systems.
• To enhance intrusion detection accuracy through the synergy of machine learning and bio-inspired algorithms.
• To introduce a new security-oriented optimization approach for interconnected networks.

Main Methods:

• Proposed a Firefly Algorithm-Machine Learning (FA-ML) technique for intrusion detection.
• Utilized a Support Vector Machine (SVM) model for classification.
• Employed the Grey Wolf Optimizer (GWO) algorithm for parameter tuning of the SVM model.
• Simulated experimental evaluations using the NSL-KDD Dataset.

Main Results:

• The FA-ML technique achieved a maximum intrusion detection accuracy of 99.34%.
• This significantly outperformed other models, with KNN-PSO achieving 96.42% and XGBoost achieving 95.36% accuracy.
• Demonstrated the effectiveness of combining machine learning with the Firefly Algorithm for robust security.

Conclusions:

• The FA-ML technique represents a significant advancement in WSN-IoT security.
• It offers a powerful and intelligent solution for bolstering intrusion detection capabilities.
• The findings validate the potential of bio-inspired algorithms and machine learning in securing modern interconnected systems.