Towards Practical Tools for Side Channel Aware Software Engineering...
summary
Presents a grey-box linear regression method that fits instruction-level power models for ARM Cortex-M0 and M4 using 5,000 traces per instruction and 1,000 traces per three-instruction sequence, sampled at 500 MS/s. Clustering reduces 21 instructions to five representatives; models include sequence-dependent interactions and adjacent-bit terms for shifts/mults. Integrates models into the ELMO emulator to simulate leakage, and reproduces a masked-AES leak confirmed by Welch t-tests with the usual 4.5 threshold.
AI-written summary · about
chapters (19)
- 0:00Introduction
- 0:46Outline
- 1:26Aim
- 2:47Userfriendly diagnostic tools
- 3:00Whitebox modeling
- 3:44Side channel modeling
- 4:24Leakage factorizations
- 6:07Linear regression
- 7:27Model quality
- 8:56Fstatistic
- 10:26ARM CortexM
- 11:15ARM Instruction Sets
- 13:07N0 Implementation
- 14:00Leakage Testing
- 15:22Grouping
- 16:14Data dependent coefficients
- 20:15Correlation
- 21:05Leakage Detection
- 23:42Conclusion
related talks
19mReality Check on Side-Channels: Lessons learnt from breaking AES on ARM Cortex-A72...
15mI Know What You Said: Unveiling Hardware Cache Side-Channels in Local Large
01h 01mTutorial: Side-Channel Attacks and Transient Execution Vulnerab...
21mA Comprehensive Framework For Power-Based Side-Channel Leakage Evaluation of SHAKTI C-Class
01h 02mCryptographic Side-Channels on Embedded Devices
36mCollide+Power: The Evolution of Software-based Power Side-Channels Attacks
42mCPU Fuzzing: Automatic Discovery of...
15mInterpretable Noninterference Measurement and its Application to Processor Designs