Outstanding StudentPoster Contest

This poster participates in

OSP

est. 2002

Departament de Matemàtica Aplicada IV

Fast-PPPassessment inEuropeanandequatorial regionnear theSolarCyclemaximum

A. Rovira-Garcia, J. M. Juan, J. Sanz

{adria.rovira, jose.miguel.juan, jaume.sanz}@upc.edu - Research group of Astronomy and GEomatics (gAGE/UPC), 08034 Barcelona, Spain

Introduction

The Fast Precise Point Positioning (Fast-PPP) is a technique to provide quickhigh-accuracy navigation with ambiguity �xing capability, thanks to an accuratemodelling of the ionosphere.

First Wide-Area Real-Time Kinematics (WARTK) feasibility studies [1] enableddi�erential relative continental navigation using only a few tens of reference stations.The further evolution [3] led to a global PPP service with the added capability of un-di�erenced ambiguity �xing, and a continental enhancement to shorten convergencebased on accurate ionospheric corrections. The technique is protected by several in-ternational patents [2] funded by the European Space Agency (ESA).

In this work, Fast-PPP is assessed in single and dual frequency users at a plane-tary scale for �rst time, and in near-maximum solar cycle conditions. This includesequatorial latitudes, more challenging than previous results, which were obtainedwith quieter ionosphere; spatial (Europe mid-latitudes) and temporal (late 2009).

Experiment Description

A world-wide distribution of rovers(green dots) is used to assess theperformance of Fast-PPP, with near-est ionospheric references located atmore than one order of magnitudegreater than typical RTK or VirtualReference Station (VRS) baselines.Other three networks are used by theCentral Processing Facility (CPF) togenerate Fast-PPP corrections; orbitand clocks, ionosphere, biases andambiguities fractional part.

Scenario DoY 150 − Year 2011

−180˚ −150˚ −120˚ −90˚ −60˚ −30˚ 0˚ 30˚ 60˚ 90˚ 120˚ 150˚ 180˚

−75˚

−60˚

−45˚

−30˚

−15˚

0˚

15˚

30˚

45˚

60˚

75˚

Global Network Continental Network Rover Network

38Fast Filter:Satellite Clocks

106Iono Filter:Regional Model

106Slow Filter: Orbits, DCBsGlobal Iono, Fract Ambigs

Satellite Orbits and Clocks

The Fast-PPP RMS orbit er-ror is maintained at the levelof 3.9 centimetres, while satel-lite clocks have an accuracyof 0.22 nanoseconds (~ 6 cm).This performance is compara-ble with International GNSSService (IGS)-Real Time com-bined product for the selectedday of the scenario is 2.7 cen-timetres and 0.21 nanosecondsfor orbits and clocks.

RMS of Precise Real-Time Satellite Orbits and Clocks wrt IGS Final

0.16

0.18

0.2

0.22

0.24

0.26

0.28

0.3

0 3 6 9 12 15 18 21 24

Clo

cks

(nan

osec

onds

)

Fast-PPP CPF Clock Corrections

0.16

0.18

0.2

0.22

0.24

0.26

0.28

0.3

0 3 6 9 12 15 18 21 24

Clo

cks

(nan

osec

onds

)

Fast-PPP CPF Clock Corrections

0.02

0.03

0.04

0.05

0.06

0.07

0.08

0 3 6 9 12 15 18 21 24

Orb

its (

met

res)

Time (Hours of Day 150 - Year 2011)

Fast-PPP CPF Orbit CorrectionsPredicted IGU Orbits

0.02

0.03

0.04

0.05

0.06

0.07

0.08

0 3 6 9 12 15 18 21 24

Orb

its (

met

res)

Time (Hours of Day 150 - Year 2011)

Fast-PPP CPF Orbit CorrectionsPredicted IGU Orbits

References

[1] M. Hernández-Pajares, M. Juan, J. Sanz, R. Orús, A García-Rodríguez, and O.L. Colombo. WideArea Real Time Kinematics with Galileo and GPS Signals. In Proceedings of the 17th InternationalTechnical Meeting of the Satellite Division of The Institute of Navigation, pages 2541�2554, LongBeach, CA, USA, September 2004.

[2] Hernández-Pajares, M., Juan, M., Sanz, J., J. Samson, and M. Tossaint. Method, Apparatus andSystem for Determining a Position of an Object Having a Global Navigation Satellite System Receiverby Processing Undi�erenced Data Like Carrier Phase Measurements and External Products Like Iono-sphere Data. international patent application PCT/EP2011/001512 (ESA ref.: ESA/PAT/566)., 2011.

[3] M. Juan, M. Hernández-Pajares, J. Sanz, P. Ramos-Bosch, A. Aragon-Angel, R. Orus, W. Ochieng,S. Feng, P. Coutinho, J. Samson, and M. Tossaint. Enhanced Precise Point Positioning for GNSSUsers. IEEE Transactions on Geoscience and Remote Sensing 10.1109/TGRS.2012.2189888, 2012.

[4] International GNSS Service Products. http://igs.org/components/prods.html, January 2014.

Ionosphere: A metric to assess ionospheric models for GNSS navigation

The core of the Fast-PPP is the capability to compute real-timeionospheric corrections with accuracies at the level or betterthan 1 Total Electron Content Unit (TECU), improving thewidely-accepted Global Ionospheric Maps (GIM), which hasnominal [4] accuracies of 2-8 TECUs. Fast-PPP large improve-ment in the modelling accuracy is achieved thanks to:

Unambiguos: Carrier-phase ambiguities are �xed in the sameCentral Processing Facility (CPF) than geodesy compu-tations (Satellite orbits and clocks).

Unbiased: Satellite DCB are estimated and removed.

Two-layer: Better separation of ionospheric and plasmas-pheric delays and its dynamics. 0

0.5

1

1.5

2

2.5

3

3.5

4

4.5

5

1 3 5 7 9 11 13 15 17 19 21 23

Pos

t-fit

Hou

rly R

MS

(T

EC

Us)

Local Time (hours of DoY 150 - Year 2011)

IGS GIM: Final combination (2h)Rapid GIM: UPC (15 min)

FPPP 1-Layer (5 min)FPPP 2-Layers (5 min)

0

0.5

1

1.5

2

2.5

3

3.5

4

4.5

5

-60 -50 -40 -30 -20 -10 0 10 20 30 40 50 60 70

24h

Pos

t-fit

RM

S (

TE

CU

s)

Geographic Latitude (degrees)

IGS GIM: Final combination (2h)Rapid GIM: UPC (15 min)

FPPP 1-Layer (5 min)FPPP 2-Layers (5 min)

A test to assess any ionospheric model for GNSS navigation is proposed:

1. Computing of Unambiguous and Unbiased (true) STEC from geometry-free combination of carrier phases measurements: STECtrue = LI− (DCBrec−DCBsat)−Bfix

I

2. RMS of the post-�t residuals of the �t of the di�erence between the STECs provided by the model under testing and the true STEC (STECtrue) to a bias per stationand per satellite: STECtrue − STECmodel = DCBrec −DCBsat

User Domain Protection Levels

A �rst glance to userdomain accuracyand its con�dencelevel is shown by theStanford plots. Thetrade-o� betweenformal and actualpositioning errors de-pend on the realismof the correctionssigmas computedby the CPF. Suchrelation has beencarefully studied toobtain safe integritymargins in the world-wide rover networkmixing several iono-spheric conditions.

All Rovers Fast−PPP 2−Freq [N=61480, 30s]

0.0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

1.0

1.1

1.2

1.3

1.4

1.5

HP

L (m

)

0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0 1.1 1.2 1.3 1.4 1.50

HPE (m)

System UnavaliableAlarm Epochs: 600 MI’

Epochs:

0

Fast−PPP

Epochs: 60880 99.02 %

95 % HPE 0.10 m

HMIEpochs: 0

MIEpochs: 0

gAGE/UPC0

1

2

3

log(n)All Rovers Fast−PPP 2−Freq [N=61480, 30s]

0.0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

1.0

1.1

1.2

1.3

1.4

1.5

VP

L (m

)

0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0 1.1 1.2 1.3 1.4 1.50

VPE (m)

System UnavaliableAlarm Epochs: 994 MI’

Epochs:

0

Fast−PPP

Epochs: 60486 98.38 %

95 % VPE 0.16 m

HMIEpochs: 0

MIEpochs: 0

gAGE/UPC0

1

2

log(n)

All Rovers Fast−PPP 1−Freq [N=61480, 30s]

0.0

0.5

1.0

1.5

2.0

2.5

3.0

3.5

4.0

4.5

5.0

HP

L (m

)

0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5 5.00 5

HPE (m)

System UnavaliableAlarm Epochs: 315

MI’Epochs:

0

Fast−PPPEpochs: 61165 99.49 %95 % HPE 0.41 m

HMIEpochs: 0

MIEpochs: 0

gAGE/UPC0

1

2

log(n)All Rovers Fast−PPP 1−Freq [N=61480, 30s]

0.0

0.5

1.0

1.5

2.0

2.5

3.0

3.5

4.0

4.5

5.0

VP

L (m

)

0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5 5.00 5

VPE (m)

System UnavaliableAlarm Epochs: 1744

MI’Epochs:

0

Fast−PPPEpochs: 59736 97.16 %95 % VPE 0.63 m

HMIEpochs: 0

MIEpochs: 0

gAGE/UPC0

1

2

log(n)

Convergence Assessment

The user state is reset every 2 hours, merging the twelve cold starts on the sameplot. Horizontal and Vertical accuracies together with its 3D formal error are plot asa funtion of time. Fast-PPP single and dual frequency positioning are respectivelyassessed against ionosphere-free strategies; GRAPHIC and classic PPP.

0 20 40 60 80 100 120Time since every 2h reset in DoY 150 - Year 2011 (min)

0.0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

Horiz

onta

l Err

or R

MS

(m)

Rover klop (8.7°E 50.0°N) 317 km to brusGRAPHICPPP 2F

IONEX 1FIONEX 2F

FPPP 1FFPPP 2F

0 20 40 60 80 100 120Time since every 2h reset in DoY 150 - Year 2011 (min)

0.0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

Vert

ical

Err

or R

MS

(m)

Rover klop (8.7°E 50.0°N) 317 km to brusGRAPHICPPP 2F

IONEX 1FIONEX 2F

FPPP 1FFPPP 2F

0 20 40 60 80 100 120Time since every 2h reset in DoY 150 - Year 2011 (min)

0.0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

Sigm

a 3D

Err

or R

MS

(m)

Rover klop (8.7°E 50.0°N) 317 km to brusGRAPHICPPP 2F

IONEX 1FIONEX 2F

FPPP 1FFPPP 2F

0 20 40 60 80 100 120Time since every 2h reset in DoY 150 - Year 2011 (min)

0.0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

Horiz

onta

l Err

or R

MS

(m)

Rover lbhu (98.7°E 0.1°N) 94 km to psmkGRAPHICPPP 2F

IONEX 1FIONEX 2F

FPPP 1FFPPP 2F

0 20 40 60 80 100 120Time since every 2h reset in DoY 150 - Year 2011 (min)

0.0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8Ve

rtic

al E

rror

RM

S (m

)Rover lbhu (98.7°E 0.1°N) 94 km to psmk

GRAPHICPPP 2F

IONEX 1FIONEX 2F

FPPP 1FFPPP 2F

0 20 40 60 80 100 120Time since every 2h reset in DoY 150 - Year 2011 (min)

0.0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

Sigm

a 3D

Err

or R

MS

(m)

Rover lbhu (98.7°E 0.1°N) 94 km to psmkGRAPHICPPP 2F

IONEX 1FIONEX 2F

FPPP 1FFPPP 2F

Conclusion

Using a world-wide distribution of rovers, the Fast-PPP technique has been assessed from the Central Processing Facility (CPF) corrections to the user domain positioning:Precise satellite orbits and clocks: are computed with an accuracy at the level to IGS real-time products; few centimetres and few tenths of nanoseconds.

Ionospheric Models are compared: Fast-PPP two-layer, ambiguity-�xed ionospheric real-time determinations better than 1 TECU (16 centimetres in L1) canbe used in combination with precise orbits and clocks maintaining their accuracy.

Reduction of convergence time: Dual-Frequency Users enhance from the best part of an hour to 5-10 minutes; in European mid-latitudes and equatorial region.

Accuraccy: The improvement of ionospheric modelling is directly translated into the accuracy of single-frequency mass-market users.

Protection levels: Safe margins are achieved thanks to the realism of the Fast-PPP corrections con�dence levels computed at the CPF level.

Acknowledgment

The authors acknowledge the use of data from the International GNSS Service, the Sumatran

GPS Array (SuGAr) network and the Asia Institute of Technology (AIT).

Funding

This work has been sponsored by the ESA Network Partner Initiative Nr. 220-211 with the

industrial partner FUGRO; Technical University of Catalonia grant FPI-UPC Nr. 120-743; Projects

GNAVIS from EC-FP7 Grant Nr. 287203 and ANTARTIDA CTM2010-21312 also contribute.