MALITE: Lightweight Malware Detection and Classification for Constrained Devices

Area of Science:

• Cybersecurity
• Machine Learning
• Computer Forensics

Background:

• Malware poses a significant threat to all computing devices, especially resource-constrained ones like IoT devices.
• Existing machine learning malware analysis methods are often too resource-intensive for these environments.

Purpose of the Study:

• To develop a lightweight malware analysis system (MALITE) suitable for resource-constrained devices.
• To accurately distinguish benign binaries from malicious ones and classify malware families.

Main Methods:

• MALITE converts binaries into images (grayscale or RGB) for analysis.
• It employs two novel, lightweight architectures: MALITE-MN (neural network) and MALITE-HRF (random forest with histogram features).

Main Results:

• MALITE-MN and MALITE-HRF demonstrate high accuracy in malware identification and classification.
• Both methods significantly reduce memory and computational resource consumption compared to state-of-the-art baselines.

Conclusions:

• MALITE offers an effective and resource-efficient solution for malware analysis on constrained devices.
• The system's low resource requirements make it ideal for mobile and IoT security applications.