IEEE Transaction on Vehicular Technology, December 2016, Vol.65, N°12
Spatial correlations found in vehicular mobility are jeopardizing the precision level of Cooperative Positioning (CP) for future Cooperative - Intelligent Transport System (C-ITS) applications. Bayesian filters traditionally assume independence of the measurement noise terms over space between different vehicles and over time at each vehicle, whereas they are actually correlated due to the local continuity of physical propagation phenomena (e.g., shadowing, multipath. . . ) under highly constrained vehicular mobility. In this paper, we break this gridlock by proposing an innovative data fusion framework capable of mitigating these effects to maintain the positioning precision level under severely correlated environments. We first illustrate the dramatic impact of correlated noise affecting both GPS and Vehicle-to-Vehicle (V2V) received power observations. Then we propose a new generic data fusion framework based on Particle Filter (PF) supporting three complementary methods to decorrelate measurement noises in a globally asynchronous context. Comparatively to conventional cooperative positioning, simulations performed in canonical vehicular scenarios (highway, urban canyon, tunnel) show that our proposed approach could provide up to 60% precision improvement in correlated environments, while matching by less than 15-20% deviation an optimal cooperative positioning scheme considered under independent measurements.
© 2016 IEEE. Personal use of this material is permitted. However, permission to reprint/republish this material for advertising or promotional purposes or for creating new collective works for resale or redistribution to servers or lists, or to reuse any copyrighted component of this work in other works must be obtained from the IEEE.