Toward Fog-Based Event-Driven Services for Internet of Vehicles: Design and Evaluation

Abstract

Internet of Vehicles (IoV) is an emerging technology for smart city. Connected vehicles can publish event data expressing their driving records so that remote vehicular cloud services can subscribe these event data and correlate them with sensed data collected from environment to provide driving services. The distributed event streams generated from heterogeneous sources can be computed for the correlation between them for on-demand traffic situation detection in IoV environment. The characteristic of event-driven service is to react to real-time service functions dependent on event trigger mechanism so that it is suitable to build situation-aware traffic applications in IoV. Meanwhile, Complex Event Processing (CEP) technology is an event stream processing technology used to compute event correlation between distributed event streams from heterogeneous sources and react to the matched specified actions immediately. CEP can be adopted to realize event-driven services in IoV. Due to the mobility of vehicles and limited bandwidth in wireless communication, the events generated from vehicles sent to remote vehicular cloud may suffer delay or get lost. Therefore, event data generated from vehicles is uncertain event due to response latency or incorrect execution of event-driven services on the cloud. Fog computing is an emerging computing paradigm that moves computation tasks from cloud to network edges and promises to reduce response latency and save bandwidth usage in wireless network for IoV. Therefore, in this paper, we propose a fog-based event-driven service mechanism for IoV and examine a case study to evaluate it.