Abstract:
This paper presents a neural-network-based SAR ADC power side-channel attack (PSA) method and a 12-bit, 1.25MS/s secure SAR ADC whose current equalizers protect the ADC f...Show MoreMetadata
Abstract:
This paper presents a neural-network-based SAR ADC power side-channel attack (PSA) method and a 12-bit, 1.25MS/s secure SAR ADC whose current equalizers protect the ADC from PSA. A prototype SAR ADC was fabricated in 65nm CMOS to demonstrate the proposed concepts. Without PSA protection, the proposed PSA method decoded the power supply current waveforms of the prototype ADC into the corresponding A/D converter output bits with >99% bit-wise accuracy except for the LSB. With PSA protection, the prototype ADC demonstrated high resistance to the proposed PSA method, showing significant drop in bit-wise accuracy.
Published in: 2020 IEEE Custom Integrated Circuits Conference (CICC)
Date of Conference: 22-25 March 2020
Date Added to IEEE Xplore: 23 April 2020
ISBN Information:
ISSN Information:
Massachusetts Institute of Technology, Cambridge, MA, USA
Massachusetts Institute of Technology, Cambridge, MA, USA
Massachusetts Institute of Technology, Cambridge, MA, USA
Massachusetts Institute of Technology, Cambridge, MA, USA
Massachusetts Institute of Technology, Cambridge, MA, USA
Massachusetts Institute of Technology, Cambridge, MA, USA