KU-UQ-AutoSec-2023
Useful links
Slide Set, Material (source code and binaries): Google Drive
Private Kaggle Competition Link: ML-based Automotive Security (Sign-up is required if you do not have your Kaggle Account.)
Tentative Course Schedule
Machine learning based automotive security
(using Cyber Security Challenge 2020 Dataset)
Monday
09:00–10:20 Introduction to CAN-based In-Vehicle Network Security & Car Hacking Dataset Overview
10:30–11:50 Conda & Pandas install
Lunch
14:00–15:20 Baseline code: Pandas usage & Data exploration
15:30–16:50 Baseline code: Time interval-based & entropy-based IDS
Tuesday
09:00–10:20 Hands-on: Build an In-vehicle IDS
10:30–11:50 Hands-on: Build an In-vehicle IDS
Lunch
14:00–15:20 Assessment (Submission through Kaggle)
15:30–16:50 Presentations from the top 3 students (based on the ML metric, F1-score)
In-vehicle infotainment system hacking
(focused on Automotive Grade Linux OS)
Thursday
09:00–10:20 Introduction (various attack sufaces on cars, how-to-hack IVI system, AGL)
10:30–11:50 AGL image install and setup
Lunch
14:00–15:20 Do your own threat analysis and write attack scenarios
15:30–16:50 Hands-on: Penetration test (target: AGL)
Friday
09:00–11:50 Hands-on: Penetration test (target: AGL)
Lunch
14:00–15:20 Assessment for the student submissions
15:30–16:50 Presentations from the top 3 students (based on Attack Score—a variation of CVSS base metric)
16:50—17:00 Conclusion Remark
lecturer
Hwejae Lee
Researcher/PhD course
Researcher/PhD course
Machine learning based automotive security (Mon–Tue)
Major Junho Jang
Pandas usage (Mon), Teaching assistant on both sessions (Mon–Tue, Thu–Fri)
