Related Experiment Videos
Implementing malware detection framework using adaptive and deformable attention-based elman residual recurrent
U Arul1, GnanaJeyaraman Rajaram1, Hari Kumar Palani2
1Department of Computer Science and Engineering, Saveetha School of Engineering, Saveetha Institute of Medical And Technical Sciences (SIMATS), Saveetha University, Chennai, Tamil Nadu 602105, India.
Computational Biology and Chemistry
|July 20, 2026
Summary
This study introduces a novel security framework for Internet of Things (IoT) networks, combining deep learning for malware detection with optimized cryptography for secure data transmission, significantly enhancing protection against cyber threats.
Area of Science:
- Cybersecurity
- Artificial Intelligence
- Cryptography
Background:
- Internet of Things (IoT) expansion introduces security vulnerabilities due to device heterogeneity and resource constraints.
- Conventional security methods are often inadequate for protecting diverse and resource-limited IoT environments.
- Malware threats pose a significant risk to the confidentiality and integrity of data in IoT networks.
Purpose of the Study:
- To design a robust and high-performance security framework for IoT networks.
- To integrate advanced deep learning for precise malware detection.
- To implement optimized cryptography for secure data transmission and efficient key generation.
Main Methods:
- Developed an Adaptive and Deformable Attention-based Elman Residual Recurrent Neural Network (ADA-El-RRNN) for malware detection.
- Utilized Opposition Humboldt Squid Optimization (OHSO) for hyperparameter tuning and efficient key generation in Optimal Key-based Elliptic Curve Cryptography (OKECC).
- Integrated ADA-El-RRNN with OKECC to create a comprehensive security framework.
Main Results:
- Achieved peak malware detection accuracy of 97.4% and specificity of 98.5% across four standard datasets.
- Demonstrated superior performance compared to existing models like LSTM, RNN, and standard ECC.
- Showcased minimized computational overhead and memory usage through OHSO-optimized OKECC.
Conclusions:
- The proposed framework offers a scalable, low-latency solution for IoT security.
- Effectively enhances data confidentiality and integrity in dynamic IoT environments.
- Provides a robust defense against malware threats, safeguarding sensitive information.