395
1 INTRODUCTION
TheionosphericeffectsonGlobalNavigationSatellite
Systems(GNSS)performanceandoperationarelong‐
standingresearchsubjectsforresearchgroupsacross
the world (Canon et al, 2013). The nature of GNSS
deterioration process due to ionospheric and space
weathereffectsdeterminestheapproachinmodelling
andmitigationofsucheffect
sincontinuityofefforts
indevelopingtheresilientGNSSandthussmoothing
andmitigatingeffectsofthefailuresanddegradations
of the GNSS performance (Thomas et al, 2011,
PetrovskiandTsujii,2012).
We conducted a study aimed at addressing the
GNSS positioning performance degradation due to
development of a space weather/ionospheric event
from the perspective of commercia
l‐grade GNSS
receivercapableofprocessingsignalsbroadcastfrom
at least to GNSS systems (the US GPS nad Russian
GLONASS), with the ability to deploy the GPS
ionosphericcorrection(Klobuchar)model(Klobuchar,
1987,SainzSubiranaetal,2013,PetrovskiandTsujii,
2012). A case‐study of a fast‐developing space
weatherdisturbanceon17t
hMarch,2015waschosen
forthisstudy.TheexperimentallycollectedGPSand
GLONASSpseudoranges datawas used as an input
for a simulation‐based study that utilised RTKLIB
(Takasu, 2013), an open‐source software‐defined
GNSS radio receiver, with capabilit
ies for post‐
processing pseudoranges with a flexibility of
modelling the positioning environment through
tailored RTKLIB settings (Takasu, 2013). Obtained
northing, easting and height time series were
analysed further with an R‐based software we
developedfortargetedpurposes.Finally,theanalysis
results were discussed and interpreted, yielding
contributiontounderstandingoftra
nsitionaleffectsof
developing space weather on GNSS positioning
performance.
The manuscript is organised as follows.
This Section presents the aim and scope of the
manuscript.
Section 2 outlines the problem and previous
research.
An SDR-based Study of Multi-GNSS Positioning
Performance During Fast-developing Space Weather
Storm
M.Filić
Sesvete,Zagreb,Croatia
R.Filjar
UniversityofRijeka,Rijeka,Croatia
L.Ruotsalainen
FinnishGeospatialResearchInstitute,Kirkkonummi,Finland
ABSTRACT:TheunderstandingoftheionosphericeffectsonGNSSpositioningperformanceformsanessential
pre‐requisiteforresilientGNSSdevelopment.Herewepresenttheresultsofastudyoftheeffectsofafast‐
developingspaceweatherdisturbanceonthepositioningperformanceofacommercial‐gradeGPS+
GLONASS
receiver. Using experimentally collected pseudoranges and the RTKLIB, an open‐source software‐defined
GNSS radio receiver operating in the simulation mode, we assessed GNSS positioning performance
degradationsforvariousmodesofGNSSSDRreceiveroperation,andidentifiedthebenefitsofutilisationof
multi‐GNSSandionosphericerrorcorrectiontechniques.
http://www.transnav.eu
the International Journal
on Marine Navigation
and Safety of Sea Transportation
Volume 10
Number 3
September 2016
DOI:10.12716/1001.10.03.03
396
Section 3 details the aim, data sources,
methodology and the software utilised for the
researchreportedinthismanuscript.
Section4presentsindetailresearchresults.
Section 5 discusses and interprets the research
results, providing the recommendation for their
practicalutilisation.
Section6concludes themanuscript
and proposes
subjectsoffurtherresearch.
2 PROBLEMDESCRIPTIONANDPREVIOUS
RESEARCH.
Global Navigation Satellite Systems (GNSS) are
essentialcomponentsofeverynationalinfrastructure,
regardless of the system ownership (Filjar, 2011).
Technologyandsocio‐economicsystemsincreasingly
rely on satellite navigation systems, and the
robustness,resilienceandreliabilityoftheir
operation
and performance (Thomas et al, 2011, Filjar and
Huljenic,2012).
Causesofsatellitenavigationmalfunctioningand
performance deteriorations are known in essence
(Sainz Subirana et al, 2013). Still, the task of
overcoming and mitigation their effects remains
unsolved, both in general and platform‐ and
application‐specific sense. (Thomas et
al, 2011)
assessedsystematicallytheriskofGNSSfailuresand
performance deteriorations, and its effects on
technology and socio‐economic systems that utilise
satellite navigation systems as enabling and
underlyingtechnology.
IonosphericeffectshaveamajorimpactonGNSS
positioningperformance among all known causes of
deterioration of GNSS positioning
performance
(Canon et al, 2013, Hapgood, 2010). Ionospheric
disturbancesresultfrommorebroaderandpowerful
processesofsolarenergy transportthroughoutspace
(Booker,1954,Davis1990,Mendillo,2006).Although
theunderstandingoftheprocessesbehindtheGNSS
ionospheric effects advances steadily (Mendillo,
2006),overcomingtheproblemremainsafar‐reaching
aim (Canon et al, 2013). The overcoming and
mitigation techniques and methods remain
application‐oriented and limited in their spatio‐
temporal outreach (Sainz Subirana et al, 2013,
PetrovskiandTsujii,2012).
The impact of developing space
weather/ionospheric disturbances on GNSS
positioning performance is even less understood,
sincethetransitional
effectsareoftenunrecorded.The
lack of experience and knowledge distilledfrom the
assessments of such events prevents successful
modelling, characterisation and correction of such
space weather/ionospheric effects on GNSS
positioningperformance.
3 GNSSPOSITIONINGPERFORMANCEDURING
DEVELOPMENTOFSTPATRICKʹSDAY2015
IONOSPHERICSTORM.
Here we address the effects
of the fast‐developing
space weather event, and the geomagnetic and
ionospheric storms, on the GNSS positioning
performance(accuracyandavailabilityofpositioning
service)usingexperimentallycollecteddatafedintoa
software‐defined GNSS radio receiver. Particular
consideration was given to assessment of potentia l
riskstomaritimenavigation.
3.1 St
PatrickʹsDay2015ionosphericstorm
A sudden solar event occurred on 17
th
March, 2015
causingseveregeomagnetic(G4,onthescaleofG1–
G5) and ionospheric storm conditions in the Earth’s
vicinity(JacobsenandAndalsyik,2016).Immediately
dubbedtheStPatrick’sDay2015storm,theeventwas
uniqueinseveralways,asfollows:
1 It was the strongest space weather event
in the
year2015.
2 International space weather watchdogs failed to
predictit.
3 Theeventcommencedsuddenly,andwiththefast
risingintensity,inthelateafternoonhours(UTC)
of 17 March, 2016, intensifying towards the
midnight.
Figure1. Planetary Kp index of geomagnetic disturbance
during the St Patrick’s Day storm event (based on data
from)
The St Patrick’s Day storm development can be
described with the time series of the planetary Kp
index of geomagnetic disturbance, as shown in Fig
1.The storm’s effects lasted for three days, affecting
numerous technology systems, including satellite
navigation.ThecharacteroftheStPatrick’sDay2015
storm offers a suitable
case for studying the GNSS
positioning performance during the fast developing
spaceweatherevent.
Figure2.GNSSSDRreceivercomponents
3.2 Datasources
The study presented here was established on post‐
processing the experimentally collected GPS and
GLONASS raw pseudoranges at the stationary
International GNSS Service (IGS, 2016) reference
station in Padua, Italy. The station’s position is
known, thus allowing for the positioning error
estimation through comparison between the
397
estimated position fixes and the known reference
pointposition.
Additionally,theIGSstandardsrequiremitigation
of all the effects of local environment, in a manner
similar to processes of the meteorological data
collection. Such a procedure exposes the space
weather/ionospheric effects and allows for their
impact on GNSS performance and
operation. The
choiceofdatasourcewasdrivenbyitsproximityto
maritime environment (the Adriatic Sea), as well as
thestation’sequipmentabilitytocollectbothGPSand
GLONASS pseudoranges. The collected daily data
sets were stored in the RINEX format and offered
freely to interested parties (scientists, researchers,
engineersetc.).TheRINEXpseudorangescollectedon
17March,2016onlywereusedtoassesstheimpactof
thestormindevelopment, particularlyin the period
fromthelateafternoontomidnight(bothUCTtime).
Figure3. RTKLIB GNSS SDR receiver setting selection
menu(settoCase2)
3.3 Softwaretools
RTKLIB, an open‐source software‐defined GNSS
receiver (Stewart et al, 2015), was used for position
estimation using raw GPS and GLONASS RINEX‐
formatted pseudoranges (Takasu, 2013).. RTKLIB
provides a complete set of the GNSS receiver
functionalities (Stewart et al, 2015), comprising
algorithmsandarchitectureforSDR
signalprocessing
and estimation at Radio Frequency (RF), baseband
andnavigation/applicationlevels(Stewartetal,2015),
as depicted in Fig 2. Since the experimentally
collected GPS and GLONASS pseudoranges were
used as input data, only the navigation (position
estimation) component RTKPOST of RTKLIB GNSS
SDRreceiverwasusedinthis
study(Takasu,2013).
3.4 Methodology
Raw GPS and GLONASS pseudoranges, along with
theinformationfromGPSandGLONASSnavigation
messages, were fed into RTKLIB, an open‐source
software‐defined GNSS receiver. With the already
identified pseudoranges at hand, only the
navigation/application component RTKPOST of the
GNSSSDRRTKLIBreceiverwas
utilised.TheRTKLIB
processing and estimation methodology and
algorithms are presented in detail in the RTKLILB
manual, provided within the RTKLIB software
package (Takasu, 2013). RTKLIB returned position
and positioning error x‐, y‐, and z‐component
estimatesatthe30s‐interval.Fivededicateduse‐cases
wereexamined,that
utilisedassistinginformationas
provided by core satellite navigation systems, GPS
andGLONASS,respectively,aspresentedinTable1.
RTKLIB/RTKPOSTwasconfiguredforaparticular
use‐case study through the RTKLIB GNSS SDR
settingselectionmenu(Takasu,2013),depictedinFig
3.Ingeneral,thesettingsresemblethoseactivatedin
commercial‐gradeGNSSreceiversdesignedforusein
maritimenavigation,thusallowingforsimulationof
therealuserGNSSequipment.
Time series of position and positioning error
estimates calculated by RTKLIB/RTKPOST were
analysed by our own software developed in the R
statistical software environment (R Development
CoreTeam,2016).Separateanalyses
wereconducted
for northing, easting and height error components,
respectively.
Table1.Use‐casedescription
_______________________________________________
Case Pseudo‐ranges Frequency Ionospheric
corrections
_______________________________________________
1 GPSSingle No
2 GPSSingle Yes
3 GPS+GLONASS Single No(GPS)
4 GPS+GLONASS Single Yes(GPS)
5 GPS+GLONASS Dual Iono‐freeLC
_______________________________________________
4 ANOUTLINEOFRESEARCHRESULTS.
Time series analysis of GNSS positioning error
componentsduring theday076in2015(17thMarch,
2015 – the day of the commencement of the St
Patrick’sDaystorm)arepresentedinthisSection.
Graphical presentations (Figs 4‐8) comprise
diagrams of the easting
(E‐W), northing (N‐S) and
height (U‐D) errors time series, respectively,
segmentedperuse‐casescenariosdepictedinTable1,
as returned by RTKPOST software component of
RTKLIB(Takasu,2013).
Figure4.GPS‐only,withoutionosphericcorrections
398
Figure5.GPS‐only,withbroadcastionosphericcorrections
Figure6.GPS/GLONASS,withoutionosphericcorrections
Figure7. GPS+GLONASS, with broadcast ionospheric
correction
Figure8.Iono‐freedouble‐frequencyGPS
Timeseriesarefurtherprocessedwithadedicated
R‐basedsoftwarepackagedevelopedbyourteamin
ordertocalculatethebasicdescriptivestatisticsofthe
timeseriesunderassessment.Thestatisticalanalysis
resultsaresummarisedinTable2.
Finally, the nature of positioning error processes
was assessed through positioning error
component
histogram analysis. Histograms of northing, easting,
andheighterrortimeseries,respectively,forthefive
use‐case scenarios were produced using R‐based
software,andarepresentedinFigs9‐13.
Figure9.Histogramsofnorthing,easting,andheightGPS‐
only uncorrected positioning errors, respectively, on 17
March,2015
Table2.GPSpositioningperformancedailystatisticsfor17
th
March,2015
__________________________________________________________________________________________________
Northingerror Eastingerror Heigherror Northingerror EastingerrorHeighterror
meanmeanmeanstandarddeviation standarddeviation standarddeviation
__________________________________________________________________________________________________
Case1 4.9413.8429.5331.1060.8242.468
Case2 3.5042.6706.4501.1080.7482.080
Case3 3.5083.0017.1120.4280.4621.099
Case4 2.6992.2215.2190.6680.5491.321
Case5 2.1701.7754.2990.3090.2600.788
__________________________________________________________________________________________________
399
Figure10. Histograms of northing, easting, and height
iono‐corrected GPS‐only positioning errors, respectively,
on17March,2015
Figure11. Histograms of northing, easting, and height
uncorrected GPS+GLONASS positioning errors,
respectively,on17March,2015
Figure12. Histograms of northing, easting, and height
iono‐corrected GPS+GLONASS positioning errors,
respectively,on17March,2015
Figure13. Histograms of northing, easting, and height
iono‐free GPS+GLONASS positioning errors, respectively,
on17March,2015
5 DISCUSSION
The analysis and interpretation of the study results
(Section4)revealsanumberofeffectsonthe GNSS
positioning performance, as follows. Those listed
below may be considered in GNSS resilience
development and the risk assessment of the GNSS
utilisationasanunderlyingandenablingtechnology,
aswellas
inmodifica tionoftheexperimentalGNSS
positioningperformanceassessmentprocedures.
5.1 Multi‐GNSSapproach reducesconsiderablystandard
deviationofpositioningerrors
The utilisation of the combined satellite navigation
systems’signals(GPS+GLONASS)slightlyimproves
the means of the positioning error components.
However,thestandarddeviationsofthepositioning
error components are
considerably improved, with
the dissipation of the positioning samples lowered
up to 50%, compared with the single‐GNSS
approach. Further to this, the time series diagrams
analysis reveals more smoothed dynamics of
positioningerrorsamples,comparedwiththecaseof
utilisationofsingle‐systempseudoranges.
5.2 Utilisationofthebroadcastparameters
oftheGPS
ionosphericmodelimprovesdailymeanpositioning
error
TheutilisationofthestandardKlobucharmodelfor
GPSionosphericcorrectionreducedbyupto30%the
positioningerror,evenwhenthecommencementofa
spaceweatherstormoccurred.
5.3 Dual‐frequencyiono‐freeGNSSpositionestimation
providesthebest
GNSSpositioningperformance
This is the fact well‐established in conclusions of
numerous research studies. Still, the navigation
market is overwhelmingly populated with single‐
frequency GNSS/GPS receivers. The GNSS
modernisation is expected to change the trend and
initiateincreasingmarketshareofmulti‐GNSSmulti‐
frequency user equipment. While this is
to happen,
theresearchersmayconsidertheutilisationofdual‐
frequency data as the reference for comparison in
studiesofGNSSperformance.
5.4 Commencement ofaspaceweatherstormcausesnon‐
GaussiandistributionsoftheGNSSpositioningerror
components
The analysis of histograms of the northing, easting
and height positioning
error components reveals
distraction from normal (Gaussian) distribution.
Whilethesimilareffectsmaybefoundinsituations
of multiple GNSS error sources in operation
(combinedinfluenceof the ionosphericstorm and a
strong urban‐area multipath, for instance), the IGS‐
baseddatausedinthisstudyassuresnoothereffects
apart
fromtheionosphericones.Thus,theconclusion
maybedrawnofthedirecteffectofthedeveloping
ionospheric storm on the daily GNSS positioning
errorstatisticaldistribution.
400
The results of the study revealed a considerable
impactonmaritimenavigationand the applications
in maritime segment. A space weather event
developmentmaycauseconsiderabledegradationof
GNSS positioning performance, but situation may
become worsened due to a more complicated
positioning error dynamics resulting from storm
development. Such effects
may affect not only the
traditionalGNSS‐basedmaritimenavigationservices
and applications, but also the emerging ones,
including: autonomous surface and underwater
vessels, automated search & rescue operations and
various robotic applications. The above‐stated
findings of this study aimed to contribute to
development of more robust and resilient
GNSS
developmentrelatedtomaritimesegment.
6 CONCLUSIONANDFUTURERESEARCH
Studies of GNSS operation and positioning
performance in situations and events of potential
GNSS disruptions create a evidence‐based
foundationfortheresilientGNSSdevelopment.Here
wepresenttheresultsofastudyofGNSSpositioning
performanceina transitional
periodofadeveloping
space weather/ionospheric event. The study was
conductedthroughdeploymentofafully‐functional
open‐source software‐defined GNSS radio receiver
RTKLIB, fed with the experimentally collected the
GPS and GLONASS pseudoranges collected
experimentallyattheIGSreferencestationinPadua,
ItalyduringaG4‐gradespace
weatherevent(storm)
in2015.
Thestudyrevealedpotentialsformitigationofthe
effects of developing space weather processes on
GNSSpositioningperformance,including:
smoothing the positioning error dynamics and
confinement of positioning error samples
dispersion through utilisation of multi‐GNSS
systems (the utilisation of satellite signals
belonging to different
satellite navigation
systems),
reductionofdailymeanpositioningerrorthrough
utilisation of the ionospheric delay correction
models,
non‐Gaussian statistical distributions of daily
positioningerrorsatthetimesofspaceweather
disturbances development, which suggests
developmentofcorrectionmodelsthatrespectthe
characterofthecausesofpositioning
errors,and
utilisation of dual‐frequency positioning error
estimates as the reference in GNSS positioning
performanceassessment.
This study addressed the case of utilisation of
GPSandGLONASS,thetwofully‐operationalglobal
navigationsatellitesystems.Futureresearchwilltake
into account effects and benefits brought by
emerging (BeiDou and
Galileo) and augmentation
(EGNOS)systems.
REFERENCE
Booker, H. G. (1954). Morphology of Ionospheric Storms.
ProcNAS,40,pp931–943.
Cannon,P etal.(2013).Extremespaceweather:impactson
engineeredsystemsandinfrastructure.RoyalAcademy
of Engineering. London, UK. Available at:
http://bit.ly/11OdBNN,accessedon10February,2016.
Davis,K. (1990).Ionospheric Radio.PeterPeregrinus Ltd.
London,
UK.
Filjar, R and Huljenić, D. (2012). The importance of
mitigation of GNSS vulnerabilities and risks.
Coordinates,8,14‐16.Availableat:http://bit.ly/Mv4QSc
,accessedon19August,2016.
Filjar,R.(2011).GNSS:Whatcangowrong,andwhatare
the risks? Risk Management in GNSS Malfunctioning
Conference.InternationalInstitute
fortheUnificationof
Private Law (UNIDROIT). Rome, Italy. Available at:
http://bit.ly/2aFS1St,accessedon19August,2016.
Hapgood, M. (2010). Lloydʹs 360º Risk Insight – Space
Weather: Its impact on Earth and implications for
business. Llyodʹs. London, UK. Available at:
http://bit.ly/1dsXhrF.
IGS. (2016). International GNSS Service. Available at:
http://www.igs.org/,accessedon:29September,2016.
Jacobsen,KS,andAndalsvik,YL.(2016).Overviewofthe
2015 St Patrick’s day storm and its consequences for
RTK and PPP positioning in Norway. J Space Weater
SpaceClim,6,A9,12pages. DOI:10.1051/swsc/2016004
Klobuchar,JA.(1987).IonosphericTime‐Delay
Algorithm
forSingle‐FrequencyGPSUsers.IEEETransonAES,3,
325–331.
Mendillo,M.(2006).Stormsintheionosphere:Patternsand
processesfortotalelectroncontent.RevGeophys,44(47
pages).doi:10.1029/2005RG000193
Petrovski, I G, and Tsujii, T. (2012). Digital Satellite
Navigation and Geophysics: A Practical Guide with
GNSS Signal
Simulator and Receiver Laboratory.
CambridgeUniversityPress.Cambridge,UK.
R Development Core Team (2016). R: A language and
environmentforstatisticalcomputing.RFoundationfor
StatisticalComputing,Vienna,Austria.ISBN3‐900051‐07‐0.
Available at: http://www.R‐project.org, accessed on 5
February,2016.
Sanz Subirana, J et al. (2013). GNSS
Data Processing –
Volume I: Fundamentals and Algorithms. ESA.
Noordwijk,TheNetherlands.
SPIDR. (2016). Space Physics Interactive Data Resource
archive. Available at: http://spidr.ionosonde.net/spidr/,
accessedon:29September,2016.
Stewart, R W et al. (2015). Software Defined Radio using
MatLab & Simulink and the RTL‐SDR. University of
Strathclyde. Glasgow, Scotland,
UK. Available at:
http://www.desktopsdr.com,accessedon29September,
2016.
Takasu, T. (2013). RTKLIB: An Open Source Program
Package for GNSS Positioning. Software and
documentation available at: http://www.rtklib.com,
accessedon19August,2016.
Thomas,Metal.(2011).GlobalNavigationSpaceSystems:
reliance and vulnerabilities. Royal Academy of
Engineering. London, UK.
Available at:
http://bit.ly/1vrIenu,accessedon19August,2016.
