PB173 Domain specific
development: side-channel
analysis
Seminar 6: Finalizing on First Steps
Łukasz Chmielewski
chmiel@fi.muni.cz, Consultation: A406 Friday 9:00-11:00
SHORT EXAMPLE
2
Guessing entropy/Key rank
3
Guessing entropy in the wild
4
Conclusion
• That is all ☺
• Last week DPA actually worked
– We looked at 0th byte instead of first
• If we have time we will look at another notebook
today (see IS)
5
ORGANIZATIONAL
6
Final Division
• Group 1: Tomas Re, Tomas Ro, Martin
– Topic: Visualization
– GitHub repository: https://github.com/reznakt/pb173-sca-visualization
• Group 2: Michael T, Lubomir, Richard
– Topic: Standard Processing, Michael might touch also “Parallel
computations with acquisition”
– The group is 3 people since Vendelín left.
– GitHub repository:
https://github.com/LubJur/PB173_standard_signal_processing
• Group 3: Michal, Matus, Filip
– Topic: Align
– GitHub repository: https://github.com/mr-akiio/trs-alignment
7
Reminder: Colloquium
• To get the colloquium
– You must be present at seminars (2 absences OK)
– You must be active at seminars (+2 points given by me at
the end)
– You must submit and get:
• 50%: 7 points in total
(projects + presentation + activity = 14 points)
9
(Modified) Seminars Plan
• 7: today, no points
• 8: evaluation of first steps given last week: 3 points per
group (personalized per person based on Github
activity) + Giving new tasks
• 9: Checking Progress: helping & trying to run your tools
• 10: 4 points per group (personalized per person based on
GitHub) + a short 5-10minuts progress presentation (1
point) + Giving new tasks
• 11: Checking Progress [Online]
• 12: Final seminar: final short 5-10minuts presentation (1
point) & grading + grading (2 points for final tasks) + 2
points for activity.
10
WHAT WAS DONE + GIVING NEW TASKS
12
Group 1: Standard Signal Processing
• First Tasks:
– Implement easy modules: average, standard deviation, histogram, absolute value,
– You can have a look at SaveAs.py and correlation.py
– Try to implement computing spectrum, some inspiration:
https://realpython.com/python-scipy-fft/
• GitHub:
13
Group 1: Main Goals
• Main Tasks:
– Spectrogram
– Incremental Correlation:
https://eprint.iacr.org/2022/253.pdf
– Pipelining
– Signal-To-Noise Ratio and other metrics
14
Group 2: Visualization
• First Tasks: implement displaying traces using 2-3
different libraries
– Matplotlib, bokeh, search for more
– Someone did some work on that. Have a look here, but it might be chaotic:
https://github.com/nilswiersma/pywf/tree/master
• GitHub:
15
Group 2: Main Goals
• I assume that displaying traces works
– Multiple traces?
• Main Tasks:
– Efficiency analysis compared to Matplotlib
– Moving traces around?
– Selecting part of the trace to run something (any code)?
– Try another library?
16
Group 3: Alignment
• First tasks:
– investigate cross-correlations in python
– See all the uploaded scripts
– Especially SaveAs.py and correlation.py
• GitHub: are disabled so I cannot see
statistics. Please enable it!
However:
17
Group 3: Main Goals
• I assume that some alignments works
• Main Tasks:
– Finish Peak-based and Correlation-Based Alignments
– Improve Efficiency
– Two from:
• Trace alignment algorithm for suppressing the clock jitter,
see pages 45-50 of: https://ged.biu-
montpellier.fr/florabium/jsp/win_main_biu.jsp?nnt=2014MON20039
&success=%2Fjsp%2Fwin_main_biu.jsp&profile=anonymous
• Elastic alignment algorithm or
• Round Based Alignment
18
Parallel computations with acquisition
• Michael
• ?
19
WALK-AROUND + GRADING
20
CHIPWHISPER?
21
Reading
• For interested people
• Side-Channel Analysis – blue book:
– http://dpabook.iaik.tugraz.at/
– The books is available at the uni.
– Look online
• The Hardware Hacking Handbook:
– https://nostarch.com/hardwarehacking
– I have an epub version.
22
23
| PA193 - Programming in the presence of side-channels /
faults
Questions