Previous Abstract Return to Session B3 Next Abstract

Session B3: Precise GNSS Positioning Applications

ECEF Position Accuracy and Reliability for Connected and Autonomous Vehicle Requirements
F. Rahman, E. Aghapour, J.A. Farrell, University of California, Riverside
Location: Cypress

The past decade has observed the rapid acceptance of GNSS navigation systems on automobiles and unmanned vehicles, especially for driver routing [1], [2]. Standard GNSS aided navigation provides good accuracy (~10 meters) which is sufficient for such applications [3], [4].
GNSS measurement errors can be categorized into two types: common mode and non-common mode [5]. Differential GNSS broadcasts and uses correction signals intended to remove the effects of the common mode errors experienced by all GNSS receivers in the same vicinity [6]. The communication channel will insert latency between the time-of-measurement and of the time-of-availability of the differential corrections.
Commonly cited position accuracy levels for pseudorange differential GNSS are 1-3 meters [7]. The lower end of this range makes GNSS systems infeasible for important safety-of-life applications if sufficient reliability is required. The FHWA, state DOT’s, and auto manufacturers are investigating connected and autonomous highway vehicle applications which will benefit from GNSS-derived, real-time, ECEF position estimates accurate to sub-meter levels at 95% probability. Pilot projects are ongoing in at least four locations [8], [9], [11], [12]. The objectives are to improve roadway network safety and throughput, while decreasing the emissions impact, through connected vehicle technologies. Navigation systems [12] achieving these accuracy and reliability specifications have not yet been demonstrated. For a national scale of implementation, topics of interest include: communication physical layers, position error sensitivity to communication latency, and positioning algorithms robust to communication latency to achieve the positioning specification.
This article will: review the tradeoffs between certain physical layer communication approaches (e.g., dedicated short-range radios (DSRC), cell towers, and satellite); present data from a study of positioning error as a function of communication latency; and discuss modeling and estimation algorithmic choices as they affect positioning algorithm performance. The objective is to achieve sub-meter accuracy with at least 95% probability. This specification will be specifically evaluated. The methods used herein will focus exclusively on the pseudorange and Doppler measurements, not carrier phase.
Novel Contribution 1: The paper will present and analyze a differential correction latency compensation approach. Previous literature discusses latency analysis and compensation methodology within the selective availability environment [13], [14]. The communicated data for the proposed approach complies with the RTCM 104 communication standard; however, the rover computations are non-standard to enhance removal of multipath and compensate for communication latency. The article will include demonstration results that maintain positioning accuracy with latencies up to 600 seconds.
Novel Contribution 2: The paper will present position estimation theory, algorithms, and experimental results that illustrate alternative modeling choices and their impact on performance and reliability. The alternative choices relate to the modeling of the multipath states, which are the dominant error source after the differential corrections remove the common-mode errors. The theoretical discussion will cover observability and the enhancement in the degree-of-observability that is achieved by modeling multipath states for each satellite and using both pseudo-range and Doppler residuals for estimation.
Novel Contribution 3: The paper will include experimental results that show the incremental improvement in positioning performance that is achieved by correction latency compensation; multi-path state augmentation; and use of Doppler to enhance the degree-of-observability. Previous literature reports position estimation enhanced performance via inclusion of Doppler measurement [15], but does not provide explanation about the observability and degree-of-observability.
Connected and autonomous vehicle applications require real-time positioning with inexpensive real-time computations. The application also needs consistent performance even in the presence of real-life inconveniences such as communication latency. The motivation of the article is to address such practical issues while reliably achieving sub-meter positioning accuracy.

Bibliography
[1] B. Hofmann-Wellenhof, H. Lichtenegger and J. Collins, “Global positioning system: theory and practice”, Springer Science & Business Media, 2012
[2] P. Misra and P. Enge "Special issue on global positioning system." Proceedings of the IEEE, vol. 87(1), pp. 3-15, 1999.
[3] G. Blewitt, "Basics of the GNSS technique: observation equations." Geodetic applications of GNSS, pp. 10-54, Nordic Geodetic Commission Sweden, 1997.
[4] C. Shuxin, Y. Wang, and C. Fei. "A study of differential GNSS positioning accuracy." In 3rd International Conference on Microwave and Millimeter Wave Technology, pp. 361-364, 2002.
[5] P. Misra and P. Enge. "Global Positioning System: signals, measurements and performance second edition." Massachusetts: Ganga-Jamuna Press, 2006.
[6] B. W. Parkinson and P. K. Enge. "Global Positioning System: Theory and applications.” vol. 2, pp. 3-50, American Institute of Aeronautics and Astronautics, 1996.
[7] P. Teunissen, and O. Montenbruck. “Springer handbook of global navigation satellite systems.” Chapter 26(2), pp. 762, Springer, 2017.
[8] “Connected Vehicle Pilot Deployment Program Phase 2, Data Privacy Plan – New York City.” [Online]. Available: https://www.its.dot.gov/pilots/cv_pubs.htm , December 2016.
[9] “Connected Vehicle Pilot Deployment Program Phase II, Data Privacy Plan – Tampa (THEA).” [Online]. Available: https://www.its.dot.gov/pilots/cv_pubs.htm , February 2017.
[10] “Connected Vehicle Pilot Deployment Program Phase 2, Data Privacy Plan, Version 2 - Wyoming.” [Online]. Available: https://www.its.dot.gov/pilots/cv_pubs.htm , April 2017.
[11] “Connected Vehicle Pilot Deployment Program Phase 2, Data Management Plan - Wyoming.” [Online]. Available: https://www.its.dot.gov/pilots/cv_pubs.htm , April 2017.
[12] J. A. Farrell, “Aided navigation: GNSS with high rate sensors.” McGraw-Hill, Inc. 2008.
[13] J. A. Farrell, M. Grewal, M. Djodot and M.Barth, “Differential GNSS with latency compensation for autonomous navigation.” In International Symposium on Intelligent Control, pp. 20-24, 1996.
[14] J. A. Farrell, M. Djodat, M. Barth and M. Grewal, “Latency compensation for differential GPS.” Navigation, vol. 44(1), pp. 99-107, 1997.
[15] M. De Agostino, M. Ambrogio, and M. Gianluca. "Doppler measurement integration for kinematic real-time GNSS positioning." Applied Geomatics, vol. 2 (4), pp. 155-162, 2010.



Previous Abstract Return to Session B3 Next Abstract