A Survey on Side-Channel-based Reverse Engineering Attacks on Deep Neural Networks

Abstract

Hardware side-channels have been exploited to leak sensitive information. With the emergence of deep learning, their hardware platforms have also been scrutinized for side-channel information leakage. It has been shown that the structure, weights, and input samples of deep neural networks (DNN) can all be the victim of reverse engineering attacks that rely on side-channel information leakage. In this paper, we survey existing work on hardware side-channel-based reverse engineering attacks on DNNs as well as the countermeasures.