Control and Management of a Connected Car Using SDN/NFV, Fog Computing and YANG data models

There are several use cases that claim the need for a connected car. Among them there is the need for connectivity between vehicles and information sources, or V2V and V2X exchanges for accident prevention. In order to cope with the need for novel applications running on top of an interconnected network, the concept of fog computing appears as a realistic solution for both intra-car and inter-car data processing and decision making. This paper describes the proposed architecture and experimental evaluation of an innovative proof-of-concept (PoC) for a connected car, modeled with YANG, which can be remotely controlled using SDN/NFV and fog computing technologies. As an example, the remote control of the car might be based on a service application running on a fog node, which can be located close to a road side unit (RSU). We also propose a fog architecture in order to enable cooperative perception between connected cars. Finally, the performance evaluation uses a RESTCONF server installed in a Raspberry Pi aboard of a small car. This server is responsible for the sensors and actuators of the car and allows for its remote control from a user terminal (e.g., a smartphone, tablet, or laptop) and through the fog node, running a control application as a service.