Artificial intelligence-driven malware detection framework for internet of things environment.

Area of Science:

• Cybersecurity
• Computer Science
• Machine Learning

Background:

• The Internet of Things (IoT) environment requires robust malware detection (MD) frameworks to safeguard sensitive data.
• Existing MD frameworks face challenges in effectively protecting IoT resources from evolving threats.

Purpose of the Study:

• To develop and evaluate an innovative image-based malware detection framework specifically for the IoT environment.
• To enhance the accuracy and efficiency of malware classification in IoT devices.

Main Methods:

• Malware binaries were converted into RGB images using image conversion and enhancement techniques.
• You Only Look Once (Yolo V7) was utilized for feature extraction from malware images.
• The DenseNet161 model was optimized using Harris Hawks optimization for image classification.

Main Results:

• The proposed framework demonstrated superior performance compared to existing MD frameworks.
• High accuracy (98.65%) and F1-score (98.5%) were achieved on the IoT malware dataset.
• Excellent performance was also observed on the Virusshare dataset with 97.3% accuracy and 96.63% F1-score.

Conclusions:

• The developed image-based MD framework is effective for protecting IoT resources.
• The framework's high accuracy and robust performance make it suitable for deployment in IoT environments.
• This approach offers a promising solution for enhancing IoT security against malware threats.