Previous Abstract Return to Session D6 Next Abstract

Session D6: Algorithms and Methods

Doppler-aided Position Estimation for HS-GNSS
Francois Vincent and Eric Chaumette, University of Toulouse, ISAE-Supaero, France
Location: Spyglass

I. EXTENDED ABSTRACT
The need for a reliable Indoor Position System (IPS) is becoming more and more pressing for a huge variety of uses, such as enhanced emergency responses or navigation aids in hospitals or malls. Many solutions have been proposed to circumvent this challenging problem. Some rely on the existence of a specific network of transmitters, such as WLAN, RFiD, light, or even ultrasound, others are based on deadreckoning and doesn’t need any beacons (IMU, video), but the best promising solutions combine the two kinds of methods [1]. Among the variety of possible solutions, extending the use of GNSS receivers to the indoor environment seems to be a solution of choice. Indeed, this solution doesn’t need the use of a costly local infrastructure and almost everyone has got a cell phone including a GNSS receiver.
Nevertheless, the GNSS systems haven’t been designed for such a use and many difficulties have to be faced. The main restricting factors are the strong attenuation due to the roof or the walls, the multipaths and the possible large differences between the satellites signals powers, known as near-far effect. However, thanks to the recent advances on the microchip processing power, new GNSS receivers that can manage the signals attenuations have emerged. These so-called HS (High
sensitivity)-GNSS receivers are based on the increase of the observation time to compensate for the propagation losses. The best way to proceed would be to increase the coherent integration time, but due to the presence of data modulation,
local clock drift and environment variations, most HS-GNSS solutions have to shrink the total observation time into shorter coherent frames to compute a sub-optimal mixed coherent and non-coherent summation. Some experiments have proven the
validity of this approach allowing to deliver a few meters precision position with C/N0 up to 20 dB lower the standard outdoor level [2].
Whereas the majority of efforts aimed at improving the Pseudo-Ranges (PR) and Doppler shifts estimations in such
a constrained environment, works concerning the position estimation algorithm are seldom. The main reason is probably linked to the feeling that the standard Weighted Least Square (WLS) iterative trilateration procedure is optimal. This idea
is shown to be exact when the observation time is short, but when it increases a substantial part of the position information, contained in the Doppler measurements is lost [3]. Indeed, two kinds of information are included in the measurement:
the Time of Arrival (ToA) contained in the PR, but also the Frequency o Arrival (FoA) contained in the Doppler shifts. It has to be noticed that some former positioning systems use to exploit this information, just like the first GNSS i.e. TRANSIT.
The information brought by the Doppler can also be viewed as an angular information leading to a classical triangulation position estimation process. Another possible interpretation of this frequency information is also the Direction of Arrival
(DoA) provided by the synthetic aperture antenna drawn from the satellites motion [4]. Whereas the ranging resolution remains fixed as the coherent integration time increases, the FoA or DoA resolution improves, making this last information
a must-use for longer observation times. It has to be noticed that this information is totally different from that exploited in the precise GNSS devices, that track the carrier phase to improving the position from an initial point. Here the Doppler information is directly used to providing an absolute receiver position.
Li et al. already proposed to exploit the Doppler information to improving the position estimation [5], but the weighting allowing to balance between the two sources of information, namely the PR and the Doppler, has to be chosen empirically. In [3], the WLS solution is shown to be optimal. We show that this procedure simply consists in a Doppler corrected version of the classical WLS algorithm based on the PR only. The main benefits of such an optimal procedure is obviously to improving the position estimation, but also to better mitigate the multipath effects, as we exploit a wider source of information.
In this paper, we will present a series of simulations illustrating, first, the improvement brought by the use of the Doppler information in an indoor environment. We will see that for observation times in the order of magnitude of 1 second (typically the observation times used in HS-GNSS) the position standard deviation can be reduced of the factor 2. We will also present simulations aimed at showing the benefits of such an algorithm in a multipath scenario.
II. CONCLUSION
This paper will present the benefits of exploiting the Doppler information to improving the position estimation in a HS-GNSS and indoor context. The proposed algorithm is very simple as it only consists in a correction of the classical WLS trilateration procedure.With the new coming GNSS waveforms providing longer data-free sequences, this HS-GNSS position estimation procedure is all the more meaningful.
REFERENCES
[1] Hui Liu, Houshang Darabi, Pat Banerjee, and Jing Liu, Survey of
Wireless Indoor Positioning Techniques and Systems, IEEE Transactions
on systems, man, and cybernetics, Part C: Applications and reviews,
VOL. 37, NO. 6, November 2007.
[2] Giovanni A. Vecchione, et al., DINGPOS, a GNSS-based Multi-sensor
Demonstrator for Indoor Navigation: Preliminary Results, Position Location
and Navigation Symposium (PLANS), 2010 IEEE/ION, 4-6 May
2010,Indian Wells, CA, USA, USA.
[3] Francois Vincent, Eric Chaumette, Christophe Charbonnieras, Jonathan
Israel, Lionel Ries, Marion Aubourg, and Franck Barbiero, Asymptotically
Efficient GNSS Trilateration, ELSEVIER Signal Processing,
Volume 133, April 2017, Pages 270277.
[4] Ali Broumandan, John Nielsen, and Gerard Lachapelle, Indoor GNSS
Signal Acquisition Performance using a Synthetic Antenna Array, IEEE
Transactions on Aerospace and Electronic Systems, Vol. 47, Issue 2,
April 2011, Pages 1337 - 1350.
[5] L. Li, J. Zhong, and M. Zhao, Doppler-aided GNSS position estimation
with weighted least squares, IEEE Transactions on Vehicular Technology,
Vol. 60, Issue 8, 2011, Pages 3615 - 3624.



Previous Abstract Return to Session D6 Next Abstract