Location patterns and predictability of large scale urban vehicular mobility

Traffic congestions become an increasingly important issue influencing people´s daily life, and the development of vehicular networks seen to be a significant technology to solve the problem of traffic congestion. In order to solve this problem by vehicular networks and intelligent transportation system, vehicular mobility is an important role in the network and system design. However, it is a still unsolved problem that is there any regularity existed in the daily random vehicular mobility. In this paper, we study the regularity that characterizes the vehicular mobility in large-scale urban environment. Despite the mess the vehicular traces appear to look like, we find certain location patterns across the whole datasets, which imply high potential predictability in vehicle location.