The accurate location of a vehicle in the road is one of the most important challenges in the automotive field. The need for accurate positioning affects several in-vehicle systems like navigators, lane departure warning systems, collision warning and other related sectors such as digital cartography suppliers. The aim of this paper is to evaluate high precision positioning systems that are able to supply an on-the-centimetre accuracy source to develop on-the-lane positioning systems and to be used in future applications as an information source for autonomous vehicles that circulate at high speeds on public roads. In this paper we have performed some on-road experiments, testing several GPS-based systems: Autonomous GPS; RTK Differential GPS with a proprietary GPS base station; RTK Differential GPS connected to the public GPS base station network of the National Geographic Institute of Spain via vehicle-to-infrastructure GPRS communications; and GPS combination with inertial measurement systems (INS) for position accuracy maintenance in degraded satellite signal reception areas. In these tests we show the validity and the comparison of these positioning systems, allowing us to navigate, in some cases, on public roads at speeds near 120 km/h and up to 100 km from the start position without any significant accuracy reduction.

After Selective Availability (SA) was turned off, the rate of change with time of the DGPS common errors (atmospheric delay, satellite orbit and clock error) became quite slow. This inevitably leads to a requirement to modify various configurations of DGPS correction message broadcasting, and reference station (RS) managers need to examine the characteristics of GPS measurement errors with SA-off. GPS error sources are temporally and spatially decorrelated, so the DGPS user position accuracy is varied by the baud-rate of the RS, the distance between the user and the station, and the noise statistics of the receiver. We identify the minimum and maximum size of correction data, interval time, the coverage range and the baud-rate that are required to maintain the existing DGPS service. Moreover, the compatibility and accuracy can be assessed to meet the users' requirements without measurements being needed. The results in this paper are used in the study and testing for the redesign of United States Coast Guard (USCG) RS. We hope that our study will be a great help in determining the flexible factors of both the RS and the user.