A Fuzzy-Based System for Assessment of Fog Computing Resources in SDN-VANETs Considering Service Migration Speed as a New Parameter

In Vehicular Ad hoc Networks (VANETs) it is essential that the services are migrated in close proximity of users, so that a continuous low latency service is always provided for time sensitive applications. Therefore, fog computing is seen as a good solution. In this paper, we propose a fuzzy-based system to assess the data processing capability of fog layer in Software Defined VANETs (SDN-VANETs). Our proposed system determines whether fog computing is appropriate and satisfies certain needs in terms of data processing. The fuzzy-based system is implemented in SDN controllers. When a vehicle needs additional resources, it can send a request to use the available resources of a fog server in its vicinity. However, for a successful data processing, the servers should meet certain requirements. The proposed system takes into consideration the time needed for sending data to the server, the load of the server and the history of previous successful tasks handled by this server. In order to support seamless service, the proposed system considers the migration speed. We evaluate the system by computer simulations. Fog layer adequacy is high when vehicle-to-server latency is low, migration speed is fast, server load is low and server history is very good.