Abstract
Several Intelligent Transportation Systems (ITS) rely on localization to enable services ranging from comfort to safety applications. Following this same idea, the use of Dedicated Short Range Communications (DSRC) devices based on the IEEE 802.11p standard, and the Vehicular Ad Hoc Networks (VANETs) paradigm, have resulted in the deployment of several protocols, services and applications that need different localization accuracy levels. The most common solution for localization is based on Global Navigation Satellite Systems (GNSS). GNSSs have problems of unavailability in dense urban areas, tunnels and multilevel roads. Moreover, GNSS have errors in the range of 5–15 meters. These unavailability and error problems are not acceptable for several VANET and ITS safety applications. In this work, we investigate and characterize the GNSS problems of error and unavailability based on datasets of real GNSS devices installed in vehicles and develop a Geodesic Dead Reckoning solution to overcome the GNSS unavailability issue.