Biserial Miyaguchi-Preneel Blockchain-Based Ruzicka-Indexed Deep Perceptive Learning for Malware Detection in IoMT.

Area of Science:

• Cybersecurity
• Machine Learning
• Blockchain Technology

Background:

• Internet of Medical Things (IoMT) devices face significant security and privacy vulnerabilities due to malware.
• Existing malware detection methods are insufficient for IoMT environments, failing to provide adequate security and privacy.
• The detection of novel malware and its variants presents an ongoing operational and research challenge.

Purpose of the Study:

• To introduce a novel technique for improved malware detection in IoMT networks.
• To enhance the accuracy of malware detection while minimizing the time consumed in the process.
• To combine the strengths of machine learning and blockchain technology for robust security.

Main Methods:

• A novel technique, biserial correlative Miyaguchi-Preneel blockchain-based Ruzicka-index deep multilayer perceptive learning (BCMPB-RIDMPL), was developed.
• The BCMPB-RIDMPL employs a deep multilayer perceptron with feature selection using point biserial correlation and Miyaguchi-Preneel cryptographic hash-based blockchain.
• The Ruzicka index is utilized to verify hash values for accurate malware classification.

Main Results:

• The proposed BCMPB-RIDMPL method demonstrated superior performance compared to existing state-of-the-art methods.
• Experiments showed significant improvements in malware detection accuracy and Matthews's correlation coefficient.
• The method effectively reduced the time required for malware detection.

Conclusions:

• The BCMPB-RIDMPL technique offers a promising solution for detecting unknown malware and its variants in IoMT.
• The integration of machine learning and blockchain technology significantly enhances security and privacy in IoMT.
• The proposed method achieves higher accuracy and efficiency in malware detection compared to current approaches.