Improving the Trust Algorithm of the C-ITS Trust-Level Concept

Description

The security of a vehicular ad-hoc network is integral to its ability to function correctly, as the topology is far more

dynamic than in traditional networks, and no central entity is needed to enable communication. If needed, vehicles

can act as relays on their own to span up a route in this fast-changing form of network. Such so-called VANets are

used in the connected mobility domain for cooperation tasks. These networks are essential for bringing across vital

information about current events like road closures or traffic hazards. Even more important than the delivery of such

information is its authenticity. Over the last years, these networks have risen in popularity, with organizations like ETSI

and the IEEE starting to standardize protocols and other mechanisms that make up the network stack. This thesis

introduces a test bed for the distributed trust algorithm by Madl based on the framework for misbehavior detection.

Additionally, it puts forward an improvement to the original concept for securing an intelligent autonomous vehicle network.

The modification is evaluated in the newly developed test bed alongside the original concept in a simulation of a European city.

The simulation evaluates that while the original proposal already fares well in everyday situations, it has problems unfolding its

functionality when presented with small traffic amounts. The resulting modification shows an improvement in the ability of

misbehavior punishment in the network, even in lower traffic volume scenarios.