Demoz Gebre-Egziabher, Assistant Professor, Aerospace Engineering & Mechanics
The project focused on developing algorithms to speed up the Time to First Fix (TTFX) of the single-frequency carrier-phase ambiguity resolution in automotive applications--in particular, applications with safety-of-life implications, such as driver-assist systems for enhancing lane-keeping performance in narrow lanes or during inclement weather when lane markings are obscured. In these applications, obstructions from buildings and highway overpasses can result in frequent GPS outages, requiring that integer ambiguities be solved often. Furthermore, since these applications have safety-of-life implications, the required level of integrity is very high. This precludes the use of currently available single-frequency carrier-phase ambiguity resolution algorithms; they cannot provide an integer fix with a high level of integrity in a short period of time. The research developed a method by which velocity and related information from sensors normally available on modern vehicles (e.g., Doppler radars, optical velocimeters) can be used to speed up the carrier-phase GPS initialization time. The approach uses velocity constraints and baseline length constraints from two or more antennas on a vehicle to speed up the TFFX given a prescribed level of integrity. Furthermore, aiding sensor accuracy requirements were analyzed to improve the robustness of the carrier-phase system after emerging from signal outages.