An in-chip countermeasure against static side-channel analysis attacks, built for deployment in FPGAs.
Static side-channel analysis exploits leakage of circuit state elements while their stored values are not changing. Typically, inducing such a state requires stopping the circuit input clock. Our countermeasure serves to monitor the clock signal and upon detecting a stop condition, immediately wipe sensitive register contents in a secure manner.
These designs are artefacts of research that has been published in a paper titled On Borrowed Time – Preventing Static Side-Channel Analysis, to appear in NDSS '25. Preprint available online here or arXiv.
This repo contains design files for two cryptographic systems that are implemented directly in hardware, each of which is equipped with the Borrowed Time countermeasure. Each instance is designed for implementation on a specific target IC since the countermeasure operates based on the physical properties of the underlying technology.
AES128
- no other countermeasures - deployment on Xilinx Kintex-7 (XC7K160T-1FBG676C)SKINNY-128-128
- first-order masking protection - deployment on Xilinx Spartan-6 (XC6SLX75-2CSG484C)
Porting these designs to other targets requires some additional engineering, namely to ensure the delay-chain circuits are correctly tuned. See paper for more information.
These applications are based on open third-party cores:
Original source:
- (
AES
core) Copyright (C) 2012, 2013 AIST - (
SKINNY
core) Copyright Corentin Verhamme and UCLouvain, 2022
Modified source - Copyright 2024 by Robbie Dumitru, Ruhr University Bochum, and The University of Adelaide, 2024
These applications can be freely modified, used, and distributed as long as the attributions to both the original author and author of modifications (and their employers) are not removed.
- an ARC Discovery Early Career Researcher Award number DE200101577
- an ARC Discovery Project number DP210102670
- the Defence Science and Technology Group (DSTG), Australia under Agreement ID10620
- the Deutsche Forschungsgemeinschaft (DFG, German Research Foundation) under Germany's Excellence Strategy - EXC 2092 CASA - 390781972