Grid-based hierarchy structure for mining and querying vehicular ad-hoc networks

Abstract

Vehicular ad-hoc networks (VANETs) are key players in the future self-organizing traffic information systems. Such systems provide safe environments on roads as they notify vehicles about congestions, accidents, and dangerous situations. VANETs facilitate broadcasting and geocasting notification messages among vehicles for the purpose of distributing important information among them. Mining and querying VANET is a key challenging task due to the spatial nature and huge amount of the information collected from the VANET's field. This paper adopts a modified version of STING (STatistical INformation Grid [1]) for building a grid-based spatial structure for storing the data collected from VANETs. The modified structure will allow an efficient querying and mining for interesting knowledge and patterns regarding vehicles and road conditions, as of the proposed framework implements a third dimension to represent time.