Series B — Session 01: Threat Modeling (Automotive IVI)
Objective - Teach lightweight threat modeling for IVI SoC: identify assets, attackers, entry points, and prioritize mitigations.
Prereqs - Pen & paper or a simple text editor; example templates provided below.
Fixed template - Assets: bootloader, kernel, userland daemons, keys, firmware images, user data - Entry points: USB, Bluetooth, Wi‑Fi, CAN bus, local USB debugging, OTA - Attackers: local attacker with physical access, remote network attacker, supply‑chain adversary
Correct Path (DIY steps)
- List assets and map to components (e.g., Bootloader→ROM/BL2, Keys→secure storage)
- Enumerate entry points per asset (e.g., OTA -> firmware image consumption)
- For each asset, enumerate threats using STRIDE (S,T,R,I,D,E)
- Assign likelihood and impact, produce a prioritized mitigation list
Break Case (single-variable change) - Omit physical access as a threat model assumption; observe that several mitigations (secure boot, tamper detection) are deprioritized and gaps appear.
Observation - Explicit attacker models change mitigation priorities; automotive systems must assume occasional physical access to IVI devices.
Why - Threat modeling aligns engineering effort with realistic attacker capabilities and system exposure.
Hard Rules - Assume worst‑practical attacker capabilities for perimeter/external interfaces. - Map crypto usage to concrete architectural bindings (boot → signatures, OTA → signatures+TLS, logging → AEAD). - Document assumptions and revalidate as system components change.
Homework - Produce a one‑page threat model for your IVI architecture listing top 5 risks and mitigations.