Performance Analysis of Code-Phase-Based Relative GPS Positioning and Its Integration With Land Vehicle’s Motion Sensors

Abstract

Relative global positioning system (GPS) positioning is used to cancel common-mode errors such as satellite/receiver clock biases and atmospheric effects. The common approach is to use differential GPS (DGPS) carrier-phase measurements to provide centimeter-meter level accuracy. However, carrier-phase-based DGPS positioning requires resolution of integer ambiguities (IA) and is sensitive to cycle-slip, which are too frequent for land-vehicle navigation. This paper investigates the feasibility of using DGPS code-phase measurements integrated with land-vehicle's motion sensors to provide highly accurate navigation system without the overhead of IA resolution or cycle-slip detection and correction, which are complex and time consuming processes. To reduce the effect of noise associated with differential code-phase measurements, a reduced set of vehicle's sensors are used in an extended Kalman filter (EKF) employing tightly coupled integration scheme (termed as EKF-DD). Owing to bias estimation of motion sensors, the proposed system flywheels through GPS outages and mitigates multipath. The performance of the proposed system was compared, using carrier-phase based reference, with two similar integration schemes employing undifferenced GPS measurements, where atmospheric effects are mitigated using either Klobuchar model (called EKF-BC) or dual frequency receivers (designated as EKF-IF). Based on three real road tests performed in challenging GPS environments and forced GPS outages, it was found that in 2-D positioning, the proposed system performed 46% superior than EKF-BC and 21% better than EKF-IF. In altitude, EKF-DD showed 66% improvement over EKF-BC and 14% over EKF-IF. In GPS outages, the overall performance of the proposed system was 21% and 10% better than EKF-BC and EKF-IF, respectively.