DSRC Security using Radio and On-Board Sensors

Anthoney Arkadie (Western State Colorado University)
Bryan Wieger (University of California, Berkeley)

Vehicular Ad Hoc Networks (VANETs) are subject to many security challenges due to the decentralized and dynamic nature of such networks. One such challenge lies in verifying the telemetry and identity of a vehicle in both the local and global perspective of the network.

This paper investigates the use of Phase Difference of Arrival to estimate the Angle of Arrival as well as using the Received Signal Strength to approximate the location of the transmitting vehicle. This information is then cross-checked with the data from the onboard sensors and the data of the transmitted data. The data from the onboard sensors, in our case a 2-D LiDAR scan, is processed through an object identification algorithm. When any of the three sets of data do not match, we consider that the vehicle may not be trustworthy.

The level of trust assessed to a transmitting vehicle is dependent on which set of data does not correspond to the communicated telemetry. If neither the radio properties or the LiDAR data match the transmitted message, we consider the transmitter as untrustworthy. The same is true if it is just the LiDAR data that does not match. If only the radio properties do not match the transmitted message, the transmitter is flagged as undetermined. If both LiDAR and radio properties match the transmitted data, the transmitter is considered trustworthy. These classifications are influenced by the high dynamic nature of channel properties and that LiDAR data is reasonably reliable.