As a recent computing paradigm, Fog Computing is opening new sets of opportunities by bringing the computational resources, applications and services closer to their consumers, making it a good foundation for the Internet of Things (IoT). As part of the IoT umbrella, Vehicular ad-hoc Networks (VANETs) offer new communication and computing opportunities that traditional paradigms, such as Cloud Computing, are unable to solve, e.g. location-awareness, latency and capacity of detecting and acting in real time to unforeseen events. As a natural computing extension, Fog Computing can be the next requirement of VANETs, enabling the deployment of mobility-related applications and services.
In this work we propose a generic architecture for the deployment of Fog Computing applications and services in a VANET environment. Moreover, we provide a proof-of-concept system to perform data analytics in a hybrid VANET/Fog environment. Our system is used by two Fog applications, one for city traffic anomaly detection, and another to estimate the bus time of arrival to feed traveller information. The reliability of such applications are validated through real mobility information from a large vehicular testbed currently deployed. The results show that using Fog Computing with a small set of recent regional data is very suitable for this type of applications, since the estimations of traffic anomalies and bus arrival times are very similar to those provided by the Cloud. Additionally, network performance results show that Fog applications can provide reliable information in a considerable shorter period of time, while at the same time they reduce the amount of traffic over the VANET backhaul infrastructure.