Artificial Neural Network-Based Mechanism to Detect Security Threats in Wireless Sensor Networks

Area of Science:

• Computer Science
• Cybersecurity
• Network Engineering

Background:

• Wireless Sensor Networks (WSNs) are critical for data collection in diverse fields but face significant security risks from routing attacks.
• Vulnerabilities in WSNs can compromise network performance, data integrity, and sensitive information.
• Existing security measures often struggle to adapt to evolving and novel routing threats.

Purpose of the Study:

• To propose and evaluate a novel method for detecting routing attacks in WSNs.
• To enhance the security and reliability of WSNs against sophisticated threats.
• To leverage artificial neural networks for intelligent threat detection in resource-constrained environments.

Main Methods:

• Utilized feed-forward artificial neural networks (ANNs) to model WSN behavior and identify routing attacks.
• Trained and tested the ANN model using the heterogeneous CICIDS2017 dataset for robustness.
• Employed the NS2 simulator for experimental validation and performance assessment.

Main Results:

• The proposed ANN-based system achieved a high detection rate of 99.21% and accuracy of 99.49%.
• Demonstrated superior performance in detecting known and novel routing attack patterns compared to existing methods.
• Effectively minimized false positive rates, enhancing detection reliability.

Conclusions:

• The developed ANN method offers a robust and effective solution for securing WSNs against routing attacks.
• This approach significantly improves the security of wireless sensor networks, protecting sensitive data.
• The study provides a reliable mechanism for secure communication in resource-constrained WSN environments.