SCAAML AES side-channel attacks tutorial
The introduction of the SCAAML (Side Channel Attacks Assisted with Machine Learning) project was at Defcon 27 (2019). The talk is recorded at the blogpost A Hacker Guide To Deep Learning Based Side Channel Attacks (talk recording and slides). Two additional blogposts are a walkthrough the code:
- Hacker’s guide to deep-learning side-channel attacks: the theory A lightweight introduction to the side-channel attacks theory.
- Hacker’s guide to deep-learning side-channel attacks: code
walkthrough
Annotation for the code available from the
scaaml_intro.
Beware that this code is the very first version and has been archived. If
you still want to run the original version you might need an older version
of TensorFlow (2.10 should work) as opposed to what is installed by the
requirements.txtfile. Or even better you might choose to use our newer GPAM model.
See code and download datasets at
Section titled “See code and download datasets at”The archived code (be prepared to use an older version of TensorFlow) can be found at our: GitHub repository.