Privacy-Preserving IDS for In-Vehicle Networks with Local Differential Privacy

Intrusion Detection Systems (IDS) for In-Vehicle Networks routinely collect and transfer data about attacks to remote servers. However, the analysis of such data enables the inference of sensitive details about the driver's identity and daily routine, violating privacy expectations. In this work, we explore the possibilities of applying Local Differential Privacy to In-Vehicle Network data and propose a new privacy-preserving IDS for In-Vehicle Networks. We have designed and conducted various experiments, with promising results, showing that useful information about detected attacks can be inferred from anonymized CAN Bus logs, while preserving privacy.