Presentation
A Computation Interleaving Countermeasure against Profiled Side Channel Attacks
DescriptionSide-channel attacks extract a secret from a computing device modeling its secret-dependent behaviour, either with a synthetic model, or profiling an instance of the design under attack.
We propose a countermeasure based on temporal interleaving, improving on the current countermeasures making profiled models of a device not portable, while also providing first-order resistance to synthetic model attacks.
Our design compares favourably with alternative solutions in the resource consumption and timing metrics.
We evaluate the security of the design with: i) profiled side-channel attacks with feature reduction with a single and multi-device model, ii) first-order synthetic model attacks, and NICV/SNR metrics.
We propose a countermeasure based on temporal interleaving, improving on the current countermeasures making profiled models of a device not portable, while also providing first-order resistance to synthetic model attacks.
Our design compares favourably with alternative solutions in the resource consumption and timing metrics.
We evaluate the security of the design with: i) profiled side-channel attacks with feature reduction with a single and multi-device model, ii) first-order synthetic model attacks, and NICV/SNR metrics.
Event Type
Work-in-Progress Poster
TimeTuesday, July 11th6:00pm - 7:00pm PDT
LocationLevel 2 Lobby
AI
Autonomous Systems
Cloud
Design
EDA
Embedded Systems
RISC-V
Security