An Automotive Cooperative Collision Avoidance Service based on Mobile Edge Computing

Even before 5G is rolled out, Mobile Edge Computing (MEC) can be considered as a key driver towards the deployment of vehicular use cases, which pose stringent latency and bandwidth requirements to the underlying Vehicle-to-Everything (V2X) communication infrastructure. In this paper, we present a MEC-enabled cooperative Collision AVoidance (CAV) service designed to anticipate the detection and localization of road hazards by extending vehicles’ perception range beyond the capabilities of their own sensors. The CAV service is a software application that runs on MEC servers allocated at the roadside and at Mobile Network Operators’ (MNO) infrastructures. The CAV service receives ETSI ITS-G5 standard-compliant messages transmitted by vehicles: periodic Cooperative Awareness Messages (CAM), which include the position, velocity and direction of the vehicle; and event-triggered Decentralized Environmental Notification Messages (DENM), which include the position of detected road hazards. The CAV service creates a distributed dynamic map using all the received information, and sends unicast messages to each vehicle with the relevant information within its collision risk area. We have implemented and validated the operation of the CAV service using vehicles’ On-Board Units (OBU) based on OpenC2X, an open-source experimental platform supporting the ETSI ITS-G5 standard.