Implementation of a MEC-based Vulnerable Road User Warning System

摘要

5G standardization is rapidly progressing and future vehicles will be connected, exchanging along the roads and on the Internet useful information for the safety of the road users. To target this goal, the Vehicle-to-Everything (V2X) standard starts to attract a lot of automobile industry and academic attention. The 5G Automotive Association categorizes a comprehensive list of connected vehicle applications, e.g. Advanced Driving Assistance (ADA) and Vulnerable Road User (VRU), that can have an implementation in the V2X environment. The expansion from Cloud to Edge Computing brings more processing power closer to the vehicles to guarantee the required latency and the need to have reduced network churn with continuous access to the Cloud. Edge Computing is addressing this paradigm shift by aiming to offer different services environment and cloud-computing capabilities within the roads infrastructure and the access network infrastructure in close proximity to vehicles and Road Side Units. This paper first overviews the V2X communication, in particular focusing on the ADA and VRU 5GAA use cases. Then it proposes a VRU warning system, which aims to provide an exchange information between the road users (e.g. pedestrian, cyclist, vehicles) about the presence of nearby entities in case of dangerous situation using commercial devices (i.e. smartphone) and standard compliant message. Finally, an experimental performance evaluation and comparison of the latency experienced by the client device considering firstly a Multi-Acces Edge Computing (MEC) based CAM Server and, then, a cloud-based CAM server are performed.