Martin Kayondo (Seoul National University), Junseung You (Seoul National University), Eunmin Kim (Seoul National University), Jiwon Seo (Dankook University), Yunheung Paek (Seoul National University)

Modern vehicles integrate Extra-Vehicle Networks (EVNs) with In-Vehicle Networks (IVNs) to support navigation, diagnostics, and over-the-air updates. This convergence introduces an EVN platform as a new source of control messages at the IVN gateway, breaking the traditional assumption that the gateway only filters traffic from simple, isolated, and implicitly trusted legacy ECUs. Instead, the EVN platform hosts a complex EVN manager with a full operating system and multiple applications, greatly enlarging the attack surface: a compromised OS or application can spoof control messages that evade gateway filtering.
We present SECV, a runtime security mechanism that enables the IVN gateway to accurately verify EVN-originated control messages even when the EVN manager is compromised. sys mediates all EVN-to-IVN traffic inside a Trusted Execution Environment (TEE), performs per-application validation, and attaches cryptographic proofs. These proofs are verified by the IVN gateway using a Hardware Security Module (HSM), providing reliable message authentication with low overhead.
SECV addresses practical challenges in TEE–HSM trust establishment, real-time mediation, and robust attribution under compromise. Implemented on an automotive SoC with ARM TrustZone and an EVITA-compliant HSM, SECV enforces strong security guarantees with only 6.5% transmission geometric mean overhead and 1.5% additional message loss during extreme communication bursts, effectively mitigating EVN-originated attacks while satisfying real-time constraints.

View More Papers

Unveiling BYOVD Threats: Malware's Use and Abuse of Kernel...

Andrea Monzani (University of Milan), Antonio Parata (University of Milan), Andrea Oliveri (EURECOM), Simone Aonzo (EURECOM), Davide Balzarotti (EURECOM), Andrea Lanzi (University of Milan)

Read More

ViGText: Deepfake Image Detection with Vision-Language Model Explanations and...

Ahmad ALBarqawi (New Jersey Institute of Technology, Newark, NJ, USA), Mahmoud Nazzal (Old Dominion University, Norfolk, VA, USA), Issa Khalil (Qatar Computing Research Institute (QCRI), HBKU, Doha, Qatar), Abdallah Khreishah (New Jersey Institute of Technology, Newark, NJ, USA), NhatHai Phan (New Jersey Institute of Technology, Newark, NJ, USA)

Read More

ACE: A Security Architecture for LLM-Integrated App Systems

Evan Li (Northeastern University), Tushin Mallick (Northeastern University), Evan Rose (Northeastern University), William Robertson (Northeastern University), Alina Oprea (Northeastern University), Cristina Nita-Rotaru (Northeastern University)

Read More