Ad Hoc Networks | 2019
Efficient data handling in vehicular micro clouds
Abstract
Abstract Wireless communication capabilities currently transform the automotive landscape. Short-range communication technologies enable a wide range of Information and Communication Technology (ICT) application for cars, drivers, and even large-scale Internet of Things (IoT) applications. Many of such applications have complex requirements in particular related to locality of data. Recently, the concept of the vehicular cloud has been proposed to address these issues, similar to what is currently investigated in the scope of Mobile Edge Computing (MEC). Forming what we call micro clouds of cars, we establish a virtual roadside infrastructure that can not only support other cars but also complex IoT applications. In this paper, we focus on data management in such micro clouds, i.e., clusters of cars organized in a hierarchical manner. Our micro clouds can provide services in their vicinity and together form macro clouds enabling more complex services and spanning entire cities. We first present an algorithm to form micro clouds at a specific geographic location using a map-based approach. Then, we develop data management services for such dynamic clusters. Concentrating on two services, namely collect data for collecting sensor data from vehicles within the micro cloud and forwarding these (possibly in aggregated form) to the macro cloud, and preserve data for keeping location-based data at the specified geo-location by continuously handing data from cars leaving to such joining the cluster. Our evaluation results clearly demonstrate the effectiveness of our approach including all the enhancements described in the paper.