Abstract
The use of computationally intensive applications requiring substantial storage resources in vehicular ad hoc networks is on the rise due to the technology shift from smart, intelligent vehicles to autonomous vehicles. Cloud computing addresses this problem to some extent by providing computational and storage facilities for the significant amount of data generated by vehicular nodes. However, it still cannot meet the stringent requirements of highly dynamic vehicular nodes in a real-time environment. The delay-sensitive applications of vehicular ad hoc networks instill the need for computing facilities in the closest possible proximity to the vehicular nodes. Therefore, fog computing has found potential in vehicular ad hoc network applications. Fog computing brings storage and computing resources closer to vehicles. Fog-based vehicular ad hoc networks can help improve the issues of first-generation vehicular ad hoc networks, such as latency, location perception, and concurrent response. Based on existing research, using fog computing decreased the delay by 50-60% and 60-70% in vehicular ad hoc network safety and entertainment and commercial applications, respectively, compared to cloud computing. An extensive study of the amalgamation of fog computing and vehicular ad hoc networks should be conducted to understand the underlying challenges of fog-based vehicular ad hoc networks. Existing surveys only include specific applications of this technology. However, we present an extensive survey focusing on the architecture of these two technologies, their integration, use cases of fog vehicular ad hoc networks, simulation tools, important performance metrics, the challenges in multiple contexts, and the gaps that need to be addressed, as well as suggest prominent journals in this field. This paper will not only impart knowledge to researchers in this area but also provide them with insights to identify the challenges and pursue their work further in fog-based vehicular ad hoc networks.