An intelligent ransomware based cyberthreat detection model using multi head attention-based recurrent neural...

Area of Science:

• Cybersecurity
• Artificial Intelligence
• Internet of Things (IoT) Security

Background:

• The proliferation of Internet of Things (IoT) devices necessitates robust security measures against evolving threats like ransomware.
• Traditional ransomware detection methods are often insufficient against advanced and adaptable ransomware attacks.
• Deep learning (DL) and machine learning (ML) offer promising avenues for enhanced ransomware detection efficacy.

Purpose of the Study:

• To propose and evaluate a novel approach for detecting and classifying ransomware attacks in IoT environments.
• To introduce the Multi-head Attention-Based Recurrent Neural Network with Enhanced Gorilla Troops Optimization for Cybersecurity Ransomware Detection (MHARNN-EGTOCRD) model.
• To enhance the accuracy and reliability of ransomware detection systems.

Main Methods:

• Data normalization using min-max transformation.
• Feature selection utilizing the Dung Beetle Optimization (DBO) model to refine datasets.
• Implementation of a Multi-head Attention-Long Short-Term Memory (MHA-LSTM) model for ransomware detection.
• Hyperparameter optimization for the MHA-LSTM model via the Enhanced Gorilla Troops Optimization (EGTO) model.

Main Results:

• The MHARNN-EGTOCRD approach demonstrated superior performance in ransomware detection.
• Experimental validation on a ransomware detection dataset yielded an accuracy of 98.53%.
• The proposed technique significantly outperformed existing models in accuracy.

Conclusions:

• The MHARNN-EGTOCRD model provides an effective and reliable mechanism for ransomware detection in IoT settings.
• Hybrid DL/ML models combined with optimization techniques show significant potential for advancing cybersecurity.
• The study highlights the importance of advanced AI techniques in combating sophisticated cyber threats.