Improving Location-based Thermal Emission Side-Channel Analysis Using Iterative Transfer Learning

2024-12-30

AI Summary

Existing single-byte modeling approaches for side-channel attacks overlook inter-byte dependencies and suffer from limited performance under small-sample conditions. To address this, we propose an iterative transfer learning framework tailored for thermal imaging and power trace analysis. This work introduces, for the first time, an iterative transfer mechanism into location-dependent thermal radiation side-channel analysis: prior-byte model parameters are reused to progressively transfer and fine-tune network weights byte-by-byte. The method jointly processes thermal and power images using a hybrid architecture combining multilayer perceptrons and convolutional neural networks, with a staged transfer strategy. Experimental results demonstrate that the framework significantly improves average attack accuracy—by up to 12.7% over baseline methods—particularly under data-scarce and low signal-to-noise ratio (SNR) thermal imaging conditions. These findings validate the effectiveness of explicitly modeling inter-byte dependencies and leveraging knowledge transfer in low-data side-channel analysis.

Technology Category

• Machine Learning: Transfer, Domain Adaptation, Multi-Task Learning
• Computer Vision: Adversarial Attacks & Robustness
• Search and Optimization: Learning to Search

Application Category

• Security and Privacy: Large-scale security measurements
• User Modeling, Personalization and Recommendation: Attacks and countermeasures in recommendation systems
• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking

Abstract

This paper proposes the use of iterative transfer learning applied to deep learning models for side-channel attacks. Currently, most of the side-channel attack methods train a model for each individual byte, without considering the correlation between bytes. However, since the models' parameters for attacking different bytes may be similar, we can leverage transfer learning, meaning that we first train the model for one of the key bytes, then use the trained model as a pretrained model for the remaining bytes. This technique can be applied iteratively, a process known as iterative transfer learning. Experimental results show that when using thermal or power consumption map images as input, and multilayer perceptron or convolutional neural network as the model, our method improves average performance, especially when the amount of data is insufficient.

Problem

• Depth Learning Model Improvement
• Side-channel Attacks
• Data Scarcity

Innovation

• Iterative Transfer Learning
• Depth Model Enhancement
• Side-channel Attacks