In recent years, the advent of deep neural networks opened new perspectives for security evaluations with side-channel analysis. Profiling attacks now benefit from capabilities offered by convolutional neural networks, such as dimensionality reduction and the inherent ability to reduce the trace desynchronization effects. These neural networks contain at least three types of layers: convolutional, pooling, and dense layers. Although the definition of pooling layers causes a large impact on neural network performance, a study on pooling hyperparameters effect on side-channel analysis is still not provided in the academic community. This paper provides extensive experimental results to demonstrate how pooling layer types and pooling stride and size affect the profiling attack performance with convolutional neural networks. Additionally, we demonstrate that pooling hyperparameters can be larger than usually used in related works and still keep good performance for profiling attacks on specific datasets.
|Name||Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)|
|Conference||satellite workshops held around the 19th International Conference on Applied Cryptography and Network Security, ACNS 2021, 3rd International Workshop on Application Intelligence and Blockchain Security, AIBlock 2021, 2nd International Workshop on Artificial Intelligence in Hardware Security, AIHWS 2021, 3rd International Workshop on Artificial Intelligence and Industrial IoT Security, AIoTS 2021, 1st International Workshop on Critical Infrastructure and Manufacturing System Security, CIMSS 2021, 3rd International Workshop on Cloud Security and Privacy, Cloud S and P 2021, 2nd International Workshop on Secure Cryptographic Implementation, SCI 2021, 2nd International Workshop on Security in Mobile Technologies, SecMT 2021, 3rd International Workshop on Security in Machine Learning and its Applications, SiMLA 2021|
|Period||21/06/21 → 24/06/21|
- Convolutional neural networks
- Deep learning
- Side-channel analysis