Cooperative relative positioning for intelligent transportation system

Abstract

In an intelligent transportation system, preventing collision accidents from happening requires to maintain a safe distance between vehicles, which is usually computed from position information. In urban areas, however, non-correlated errors due to multipath propagation in absolute positions degrade the accuracy of relative position. On the other hand, simply removing reflected signals might lead to a shortage of satellites in fixing positions. In this paper, we suggest a cooperative relative positioning (CoRelPos) scheme. Vehicles learn measured information of nearby vehicles by inter-vehicle communications. Correlated information, including that of reflected signals, is used to compute relative position. Statistical analysis of experimental logs verifies that information measured by nearby vehicles is highly correlated under most cases. But exceptions do exist, which require correlation detection. Simulation evaluation and initial experimental results confirm that the proposed scheme can effectively improve the accuracy of relative position compared with the state-of-the-art schemes.