Efficient Selection of VANET Messages for Series Vehicles

In Vehicular Ad-hoc Networks (VANETs) communication among vehicles enables new advanced driver assistance systems. Cooperative Awareness Messages (CAMs) are so frequently exchanged that deployable electronic control units will not be powerful enough to process them all. Thus, most relevant CAMs need to be selected for processing. In this paper, we give an overview of relevance estimation function (REFs) for CAMs that assign them a relevance value based on metadata. We describe and evaluate an efficient message buffer and selection mechanism for most relevant CAMs. It efficiently decreases the relevance of buffered CAMs over time. We evaluate the results of the mechanism with regards to selection probability and message waiting time under realistic conditions. Furthermore, we study the runtime of various REFs using a prototypical implementation on a hardware evaluation platform. The results show that the proposed algorithms are feasible on close-to-production hardware as they quickly process most relevant CAMs.