Security - Nana-Kusa/can-invader-protection GitHub Wiki

Security Logic

This page provides insight into how the system defends against unauthorized CAN bus access.

Threat Model

  • Attacker injects fake "unlock" messages via OBD-II or CAN wiring.
  • Victim vehicle unlocks under assumption of keyless entry.

Defense Strategy

  • Monitor for specific unlock CAN IDs (configurable).
  • Automatically reply with a lock signal when unauthorized unlock is detected.
  • Optional: delay between unlock and lock to avoid repeated override.

Limitations

  • Cannot distinguish legitimate unlock from attacker without user presence detection.
  • Assumes attacker does not simultaneously jam response signal.