Evaluating the V2X latency for vehicle positioning: a comparison between 5G-V2X and ITS-G5

Connected and Automated Vehicles (CAVs) are revolutionizing transportation, promising enhanced safety, efficiency and sustainability. One key requirement for their safe circulation on road is related to the positioning information, which needs to be always available, with low latency and high accuracy. This work considers an infrastructure-based vehicle positioning system and it analyzes the latency of sending the position information to the vehicle with a Vehicle-to-Everything (V2X) connectivity. We measure the Round Trip Time (RTT) between a Road Side Unit (RSU) and a CAV, comparing the performances of 5G-V2X and ITS-G5 technologies. An on-field experimentation is analyzed for the former, while a laboratory environment is used for the latter. We also provide a performance comparison between a cloud and an on-premise Message Queueing Telemetry Transport (MQTT) broker, showing the latter's positive impact in reducing the V2X latency. Indeed, a cloud architecture is not a viable solution for low-latency services (RTT with peaks around 700 ms), while an on-premise broker has an RTT in the order of tens of ms, with values below 10 ms for 80% of cases, thus enabling demanding services such as infrastructure-assisted navigation.