INTERSTICE – INTelligent sEcuRity SoluTIons for Connected vEhicles

Purpose and goal

The evolving landscape of connected vehicles introduces significant cybersecurity challenges. This project focuses on developing architecture-aware distributed machine learning-driven systems for onboard intrusion detection (IDS) for vehicle communication networks, i.e., CAN and Automotive Ethernet. Network IDS provides a layer of security by monitoring and analyzing the data traffic, and identifying suspicious activities that could indicate an intrusion. It can facilitate the timely detection of threats and enable the application of appropriate mitigation measures.

Expected effects and result

The project introduces (a) innovative methods for generating attack data using advanced simulation environments and generative AI models, (b) tools and methods for analyzing network data as well as solutions for complex attack detection on CAN and Automotive Ethernet based on edge-deployable ML models augmented with explainability, (c) efficient strategies for the deployment of lightweight IDS on the edge node, such as vehicle ECUs, considering real requirements and constraints; and federated learning-based solutions to develop the vehicle´s aggregated IDS model.

Planned approach and implementation

The project is a collaboration between Scania CV AB, RISE, and Scaleout Systems AB and contains five work packages (WP): WP1 focuses on data instrumentation and preparation. WP2 addresses strategies for implementing lightweight and trustworthy ML-driven IDS for CAN networks. WP3 involves the deployment of lightweight IDS and the development of an aggregated IDS model using federated learning. WP4 focuses on IDS for Automotive Ethernet and WP5 is dedicated to the dissemination of the results.