Assessment of node trustworthiness in VANETs using data plausibility checks with particle filters

In Vehicular Ad-Hoc Networks (VANETs), the exchange of location data (i.e. absolute position, heading, time) for traffic safety applications plays an important role. The trustworthiness of this information is crucial as false data affects applications heavily and might endanger human lives. Beside cryptographic solutions that ensure sender authenticity and message integrity, the data plausibility check is an important mechanism to ensure positional reliability. In this paper, we show that a particle filter is an appropriate instrument to perform plausibility checks in order to assess the trustworthiness of neighbor nodes. Our approach allows the aggregation of information from different data sources directly in one particle filter per neighbor. Thus, dependencies and relationships between individual sources can be fully accounted for and the framework is easily extensible and scales well. The concept is implemented as a Java-OSGi bundle for a field operational test framework and evaluated using both manually generated traces and recorded data from real vehicle trips. We show that the detection of several types of location-based attacks is possible under consideration of errors and system inherent deviations in sensor data.