An Empirical Study on the Temporal Structural Characteristics of VANETs on a Taxi GPS Dataset

In recent years, vehicular ad hoc networks (VANETs) have received significant interests from both academia and industry because of the ubiquitous communication capabilities they provide. VANETs are expected to support a wide range of applications that may influence our daily life, ranging from road safety applications to entertainment ones. A key to the development of protocols and algorithms for intervehicle communication and services lies in the knowledge of the structural properties of the VANET communication graph. Nonetheless, previous researches represented the VANET communication graph as a set of snapshots and tracked the evolution of some metrics on snapshots at certain time intervals. They neglected to take into account the temporal property of VANETs, which in many real scenarios plays a pivotal role. In this paper, we mainly explore the temporal structural characteristics of VANETs. A time-extended model is first presented to capture the temporal property of VANETs. Based on the model, some key metrics are defined to describe the structural characteristics of VANETs in various aspects. We then employ a real and large-scale urban taxi GPS dataset to provide a comprehensive study of the temporal structural characteristics involved with time-ordered paths, reachability, and connectivity of VANETs. These results are helpful in designing better protocols and algorithms to achieve reliable and low-latency communications.