Positioning Enhancement with Double Differencing and DSRC

Global Navigation Satellite Systems (GNSS) are comprehensively used for navigation in vehicular environments. However, the limited accuracy of GNSS for civilian use makes it unsuitable for some safety applications such as collision avoidance. Cooperative positioning in vehicular networks is a relatively new concept for positioning enhancement in a group of vehicles capable of communicating with each other. In addition to data communication, estimation of the distance between the nodes of a vehicular network is one the most challenging issues in cooperative positioning. Radio ranging methods such as Received Signal Strength (RSS), Time of Arrival (TOA), and Time Difference of Arrival (TDOA) are the most common techniques which are referred to in the literature for distance estimation without considering the poor accuracy of these methods for positioning accuracy improvement purposes, especially in vehicular environments. In this article, two methods for positioning enhancement and distance estimation are proposed that do not rely on common radio ranging techniques. The independence from radio ranging is one of the main contributions of this work. The proposed techniques are solely based on communicating data among the vehicular network nodes through Dedicated Short Range Communication (DSRC). The proposed solutions are suitable for suburbs and open space areas such as highways with less multipath error and at least four common visible satellites. The key idea is elimination of common errors of pseudorange estimates among the nodes of the vehicular network in a cluster of vehicles or between vehicles and roadside units. This looks similar to Differential GPS (DGPS) but the approach is different. The other advantage of the proposed methods is the low minimum possible number of participating nodes required by the algorithms, which is useful for saving communications bandwidth. Some experiments were conducted to verify the performance of the proposed algorithms. Results for the two methods, tested in static situations, show gains of 50%-65% for distance estimation depending on the level of multipath error and 55% for position estimation in a low multipath environment.