Related Experiment Video
Updated: May 2, 2026
08:20
Author Spotlight: AI-Driven Trypanosome Species Detection from Microscopic Images
Published on: October 27, 2023
2.9K
Revolutionary hybrid ensembled deep learning model for accurate and robust side-channel attack detection in cloud
C Lakshminatha Reddy1, K Malathi2
1Saveetha School of Engineering, Saveetha Institute of Medical and Technical Sciences, Chennai, India. laxminathreddy842@gmail.com.
Scientific Reports
|September 26, 2025
Summary
This study introduces a hybrid deep learning model for detecting side-channel attacks (SCAs) in cloud environments. The model achieves high accuracy, enhancing cryptographic security against physical data leakage in shared computing systems.
Area of Science:
- Cybersecurity
- Cryptography
- Cloud Computing
Background:
- Side-channel attacks (SCAs) exploit physical data leakages, posing a threat to cryptographic systems.
- Detecting SCAs in cloud environments is challenging due to noise and complex data patterns.
Purpose of the Study:
- Develop a robust deep learning model for SCA detection in cloud environments.
- Ensure scalability and accuracy in identifying physical data leakages.
Main Methods:
- Proposed a hybrid ensembled deep learning (HEDL) model integrating CNN, LSTM, and AutoEncoders.
- Incorporated an attention mechanism to focus on critical data segments.
- Trained and evaluated the model on the ASCAD dataset in a cloud environment.
Main Results:
- Achieved 98.65% detection accuracy, outperforming traditional and standalone deep learning models.
- Demonstrated superior performance in both clean and noisy data conditions.
- Attention mechanism improved focus, reduced computational demands, and enhanced precision.
Conclusions:
- The HEDL model offers robust and accurate SCA detection in noisy cloud environments.
- Represents a significant advancement in cloud-based cryptographic security.
- Highlights the effectiveness of hybrid deep learning with attention for SCAs.