Lower Voltage for Higher Security: Using Voltage Overscaling to Secure Deep Neural Networks

Shohidul Islam (1) , Ihsen Alouani (2, 3) , Khaled Khasawneh (1)

1 George Mason University [Fairfax]
2 IEMN - Institut d’Électronique, de Microélectronique et de Nanotechnologie - UMR 8520
3 COMNUM - IEMN - COMmunications NUMériques - IEMN

Deep neural networks (DNNs) are shown to be vulnerable to adversarial attacks-- carefully crafted additive noise that undermines DNNs integrity. Previously proposed defenses against these attacks require substantial overheads, making it challenging to deploy these solutions in power and computational resource-constrained devices, such as embedded systems and the Edge. In this paper, we explore the use of voltage overscaling (VOS) as a lightweight defense against adversarial attacks. Specifically, we exploit the stochastic timing violations of VOS to implement a moving-target defense for DNNs. Our experimental results demonstrate that VOS guarantees effective defense against different attack methods, does not require any software/hardware modifications, and offers a by-product reduction in power consumption.

Mots clés: Adversarial attacks; machine learning; approx- imate computing; voltage overscalin

Domaines: Sciences de l'ingénieur [physics]

Shohidul Islam, Ihsen Alouani, Khaled Khasawneh. Lower Voltage for Higher Security: Using Voltage Overscaling to Secure Deep Neural Networks. 2021 IEEE/ACM International Conference On Computer Aided Design (ICCAD), Nov 2021, Munich, Germany. pp.1-9, ⟨10.1109/ICCAD51958.2021.9643551⟩. ⟨hal-03584349⟩