A GPS Velocity Sensor: How Accurate Can It Be? … A First Look

Luis Serrano, Donghyun Kim, and Richard B. Langley, Kenji Itani and Mami Ueno

Abstract:	To obtain accurate velocity information, traditional optical velocity sensors and RTK-GPS positioning are common but expensive solutions. However, velocity information can be obtained from the time-differential method using a stand-alone single-frequency receiver without resolving carrier phase ambiguities. In dynamic vehicle testing, such a GPS sensor is attractive because it enables a single person to conduct performance tests quickly and easily. We have investigated the feasibility of a low-cost GPS velocity sensor for applications such as vehicle testing. We developed post-processing software which uses the first order central difference approximation of the carrierphase rate. The advantage of this approximation is simplicity, which facilitates implementation of algorithms into the receiver. In this paper, we mainly focus on the scientific aspects of the GPS velocity determination. We investigated the potential accuracy which can be achieved with low-cost receivers, and evaluated the error budget present in the estimation. The algorithms consider specifically data smoothing, multipath, and modeling of ionospheric effects as well as efficient handling of cycle slips and other data anomalies. We conducted field tests to evaluate the performance of Furuno's low-cost single-frequency GPS receiver for a static and a kinematic case. In static mode, the user velocity has been estimated to be better than 1 cm/s (2- sigma) in a high-multipath environment. In kinematic mode, we observed the increase of receiver dynamics in the residuals. However, we did not experience any long signal interruption even under high multipath. Using the Doppler measurements (either the receivergenerated Doppler or the carrier-phase derived Doppler) observed from a moving platform, it is easy to determine the velocity of the platform as long as the satellite velocity is precisely known. From our early results, we confirmed that the satellite velocity predicted by using the broadcast ephemeris in the navigation message is comparable to the velocity of NGA (National Geospatial Intelligence agency) SP3 precise ephemeris. The errors in the position of a moving vehicle cause an error in the calculation of radial velocity. For stand-alone velocity determination, in order to achieve a solution at the mm/s level, satellite positions have to be known better than 10 m. This paper describes the algorithm developed at UNB, results and analysis of the field tests and future work to be done.
Published in:	Proceedings of the 2004 National Technical Meeting of The Institute of Navigation January 26 - 28, 2004 The Catamaran Resort Hotel San Diego, CA
Pages:	875 - 885
Cite this article:	Serrano, Luis, Kim, Donghyun, Langley, Richard B., Itani, Kenji, Ueno, Mami, "A GPS Velocity Sensor: How Accurate Can It Be? … A First Look," Proceedings of the 2004 National Technical Meeting of The Institute of Navigation, San Diego, CA, January 2004, pp. 875-885.
Full Paper:	ION Members/Non-Members: 1 Download Credit Sign In