A Data Generation Workflow for Consensus-Based Connected Vehicle Security

Due to their scalability and potential to integrate into existing communication protocols, consensus-based trust algorithms are a promising approach for detecting data falsification attacks in connected vehicle networks. Despite this potential, the standardized data sets necessary to support the development of these algorithms are not readily available. This manuscript demonstrates progress towards addressing this limitation by developing a synthetic data generation framework which — 1) simulates the generation of peer-estimated basic safety messages (BSM) within a real data set of BSM broadcasts, and 2) simulates the presence of malicious nodes within the network by injecting anomalies into peer-estimated and self-reported BSM signals.