RESEARCH ARTICLE

IPWV Estimation from Different GNSS Antenna

Ram Kumar Giri

Received: 19 July 2010 /Accepted: 31 August 2011 /Published online: 14 October 2011# Indian Society of Remote Sensing 2011

Abstract Global Navigation Satellite System (GNSS)is widely used nowadays in variety of applications.The observation file for the near real timeestimation of Integrated Precipitable Water Vapour(IPWV) received at the ground-based receiver ismixed with ambiguities. Multi-path effects affectthe positional accuracy as well as range fromsatellite to ground based receiver of the system.The designing of the antenna suppress the effect ofmulti-path, cycle slips, number of observations,and signal strength and data gaps within the datastreams. This paper presents the preliminary dataquality control findings of the Patch antenna(LeicaX1202), 3D Choke ring antenna (LeicaAR25GNSS) and Trimble Zephyr antenna (TRM39105.00). The results shows that choke ringantenna have least gaps in the data, cycle slipsand multi-path effects along with improvement inIPWV. The signal strength and the number ofobservations are more in case 3D choke ringantenna.

Keywords GNSS . Cycle slip . Signal to noise ratio(SNR) . Choke ring antenna .Multi-path

Introduction

An antenna’s job is to capture some of the power inthe electromagnetic waves it receives and to convertit into an electrical current that can be processed bythe receiver. With very strong signals at lowerfrequencies, almost any kind of antenna will do. Ingeneral, an antenna must be designed for theparticular signals to be intercepted, with the centerfrequency, bandwidth, and polarization of the signalsbeing important parameters in the design. As therequired receiver position fix accuracy approachescentimeter and even sub-centimeter levels, thedemands on the antenna increase, with multi-pathsuppression and phase-center stability becomingimportant characteristics. Various properties andtrade-offs that affect functionality and performanceof the antenna are different for different antenna.For the precise estimation of Integrated PrecipitableWater Vapour (IPWV) both the Global NavigationSatellite Systems (GNSS) receiver and antenna areequally important. The coverage of the frequenciesused by these systems are unique, such as Galileo’sE6 band and the GLONASS L1 band, and may notbe covered by all antennas. GPS satellites transmitthe radio signal, which travels into the atmospherewith different layers in it. The troposphere part ismainly responsible for weather and satellite signalwill passes in layers of varying refractive therefractivity profile can be transformed to profiles

J Indian Soc Remote Sens (September 2012) 40(3):389–398DOI 10.1007/s12524-011-0159-2

R. K. Giri (*)India Meteorological Department,Lodi Road,New Delhi 3, Indiae-mail: rkgiri_ccs@rediffmail.com

of tropospheric humidity by using the thermody-namics relations.

In the existing operationally running system signalemanating from the GPS satellites is captured byLeica X1202 (Patch antenna) and received in LeicaGRX 1200 receiver. The RINEX hourly data fromfive GPS stations (Delhi, Mumbai, Kolkata, Chennaiand Guwahati) is processing near real time basiscentrally at Delhi. Daily at each site Leica binaryobservation files are stored both on the receiver-internal Compact flash card (in ring buffer mode) andexternally on the data logging PC. The Leica binaryobservation files of the logging PC are automaticallyconverted into Receiver Independent Exchange(RINEX) format through spider software. The RINEXzip file contains three files namely; observation,navigation (broadcast ephemeris) and meteorological

data, archived and final product Integrated Precipita-ble Water Vapour (IPWV) is uploaded to IndiaMeteorological Department (IMD) web site andmade available to the end users. The nearly realtime and post processing of the observational datais also done by the GAMIT 10.3.2.1 processingsoftware. Leica GRX 1200 is a dual frequency(with fully independent L1 and L2 tracking loops)receiver. In general, the functionality and perfor-mance of the GNSS antenna depends mainly on thefollowing properties like, Frequency coverage, gainpattern, circular polarization, multipath suppression,phase center, impact on receiver sensitivity andinterference handling.

Multi-constellation bands compatibility and itscompliant with all requirements of the GNSS reviverscreates the antenna becomes harder to design.

Satellitevehicle (SV)

Av Elev (degree) MP1 (mm) MP2 (mm) Average signalstrength (SNR)a

Average signalstrength (SNR)a

L1 in dB L2 in dB

G3 11.7 0.14 0.25 4.2 7.9

G6 14.1 0.18 0.17 4.5 6.8

G9 09.4 0.24 0.30 6.7 6.8

G14 63.1 0.13 0.08 9.0 8.5

G18 33.0 0.22 0.23 8.6 1.5

G19 17.7 0.22 0.20 5.7 2.3

G21 12.3 0.25 0.27 4.7 5.8

G22 56.9 0.11 0.08 1.3 8.9

G26 28.4 0.14 0.13 8.0 2.7

G31 36.6 0.20 0.14 8.4 5.1

Table 1 Code multi-pathRMS by satellite (+) andaverage signal strength byPRN (*) for Leica Chokering (LeicaAR25 GNSS)antenna

a Signal strength is mappedbetween 1 (worst) and 9(best), 5 is the threshold forgood SNR, 0 indicates valueis unknown

Satellitevehicle (SV)

Av Elev (degree) MP1 (mm) MP2 (mm) Average signalstrength (SNR)

Average signalstrength (SNR)

L1 in dB L2 in dB

G3 11.7 0.00 0.56 40.2 22.0

G6 14.1 0.52 0.45 40.1 22.6

G9 09.4 0.35 0.00 41.2 23.8

G14 63.1 0.20 0.12 35.1 17.4

G18 33.0 0.20 0.47 40.2 23.1

G19 17.7 0.21 0.34 49.6 39.1

G21 15.3 0.38 0.00 42.5 23.6

G22 56.9 0.20 0.14 51.6 40.7

G26 28.4 0.35 0.35 45.1 31.5

G31 36.6 0.48 0.21 47.3 34.2

Table 2 Code multi-pathRMS by satellite (+) andaverage signal strength byPRN (*) for Trimble Zephyrantenna (TRM 39105.00)

390 J Indian Soc Remote Sens (September 2012) 40(3):389–398

Data and Methodology

For this present study the RINEX format data for thethree type of antenna have been taken from IndiaMeteorological Department, Lodi road New Delhi.The three antenna systems are Patch antenna (Lei-caX1202), 3D Choke ring antenna (LeicaAR25 GNSS)and Trimble Zephyr antenna, (TRM 39105.00), whosedata have been used. The antennas have been installedat Mausam Bhawan, Lodi Road, New Delhi for almost20 days (07-07-2009 to 25-07-2009) and that RINEXdata are utilized for final processing of IPWV. Thirdsystem Trimble Zephyr data could not be processed asa final output of IPWV due to system integrationfailure. The Leica GNSS QC V 2.1.2 software (Esteyand Meertens 1999) was used for quality controlanalysis of Trimble Zephyr and LeicaAR25 antennasobservations files. The observation data received from

patch and choke ring antenna is processed indepen-dently through GAMIT 10.3.2 software for IPWVestimation. For the existing old LeicaAX1202, TEQCtool kit software developed by University NAVSTARConsortium (UNAVCO) Boulder, USA has beenused. For comparison of data quality we have takenthe data of 29-07-09 (0530 UTC to 0620 UTC) for allthe three systems.

Results and Discussion

The preliminary results of the data quality obtained byusing the above said software for Patch antenna(LeicaX1202), Choke ring antenna (LeicaAR25GNSS) and Trimble Zephyr antenna (TRM39105.00) are summarized in the Tables 1, 2, 3 and4. Various parameters of data quality like multi-path,

Satellitevehicle (SV)

Av Elev (degree) MP1 (mm) MP2 (mm) Average signalstrength (SNR)

Average signalstrength (SNR)

L1 in dB L2 in dB

G3 11.7 0.00 0.56 49.33 43.32

G6 14.1 0.52 0.45 49.63 42.80

G9 09.4 0.35 0.00 49.16 42.02

G14 63.1 0.20 0.12 48.83 39.89

G18 33.0 0.20 0.47 47.89 38.22

G19 17.7 0.21 0.34 40.30 36.26

G21 15.3 0.38 0.00 35.72 32.58

G22 56.9 0.20 0.14 44.26 30.59

G26 28.4 0.35 0.35 46.29 33.21

G31 36.6 0.48 0.21 37.43 41.27

Table 3 Code multi-pathRMS by satellite (+) andaverage signal strength byPRN (*) for Patch antenna(LeicaX1202)

Av average, IOD iono-spheric delay, Num number,Obs observations, MP mul-tipath, SV satellite vehicle,Slips cycle slips

Name of the antenna→ 3D- Choke ring antenna(LeicaAR25 GNSS)

Trimble Zephyr antenna(TRM 39105.00)

Patch antenna(LeicaX1202)↓Parameters

Data gaps (time) 0.0 (5 s)* 3.0 min (5 s)* 4.0 min (5 s)*

Average of PDOP 4.8 4.8 5.8

Average of GDOP 5.7 5.7 6.2

SV (average number) 7.2 7.2 5.4

MP1 RMS (m) 0.16 (0.5)* 0.26 0.26

MP2 RMS (m) 0.17 (0.5)* 0.24 0.38

Total cycle Slips (number) 0 (2)* 11 (2)* 12

Clock offsets (millisecond) 0.00025 0.00019 0.00012

L1 (SNR) (average) dB 7.0 44.9 47.7

L2 (SNR) (average) dB 5.4 29.7 39.5

EPOCH 100% (99)* 95.8% (99)* 95.2% (99)*

Table 4 Comparison ofdifferent parameters ofantennas

Values in the above table arethreshold values to passthe quality test

*Represents the permissiblestrength of the antennaparameters

J Indian Soc Remote Sens (September 2012) 40(3):389–398 391

cycle slips and SNR values for all the three systems aregiven in Table 4. Antenna design plays an importantrole in minimizing the errors in the estimation ofIPWV (Figs. 1 and 2). The processed IPWV withactual observations for choke ring and patch antenna areshown in Figs. 3 and 4. The various graphical plots forsignal strengths and Signal to Noise Ratio (SNR) forboth the frequencies (L1 & L2) are shown in Figs. 3, 4,5, 6, 7, 8, 9, 10, 11 and 12. The plots from Figs. 5, 6, 7and 8 are for Leica 3D Choke ring AR25 GNSS

antenna and plots from Figs. 9, 10, 11 and 12 are forTrimble Zephyr antenna, 2D antenna. All antennas areomni-directional. It has been clear from the observedanalysis of the data that the data gaps, cycle slips areleast in Leica 3D Choke ring antenna.

Leica Choke Ring Antenna Axial Ratio 25 (AR25)

AR25 choke ring geodetic antenna covers the entirefrequency range for GPS L1, L2 and L5, GLONASSL1, L2, GALILEO L1, E5, E6 bands. The use of chokering ground plane is typical in geodetic antennas. Theseallow good gain pattern control, excellent multi-pathsuppression, high front-to-back ratio and good AR atlow elevation angles. Choke rings contribute to a stablephase center. Leica AR25 is the best usable antenna forthe high-end geodetic applications.

Multi-path Effect

Multi-path reflections from ground (Fig. 1) and fromvertical can introduce (Fig. 2) extra path length andintroduce an error in the actual estimation of rangefrom the GPS satellites. Positional accuracy isimproved by using the choke ring antenna. Multi-path susceptibility of an antenna can be quantifiedwith respect to the antenna’s gain pattern character-istics by the multi-path ratio (MPR). MPR for groundreflections is given below:

MPR ¼ ERHCP qð ÞERHCP 180� � qð Þ þ ELHCP 180� � qð Þ ð1Þ

In the above Eq. 1 the reflected signals have moreLeft Handed Circularly Polarized (LHCP) componentthan Right Handed Circularly Polarized Light(RHCP). In this case the antenna gain is suitablyadjusted with the help of MPR and by making theproper design of antenna (Moernaut and Orban 2009).

Similarly the vertical reflections are minimized bysuppressing the reflections of multiple objects bysuitably adjusted the MPR.

MPR ¼ ERHCP qð ÞERHCP qð Þ þ ELHCP qð Þ ð2Þ

The typical multi-path error in the pseudorangemeasurement varies from about 1 m to 5 m. Multi-path in the pseudorange measurements are usuallytwo orders of magnitude bigger than the multi-pathFig. 2 Reflections from vertical

Fig. 1 Reflections from ground

392 J Indian Soc Remote Sens (September 2012) 40(3):389–398

errors in the carrier-phase measurements (Langley1998; Misra and Enge 2001)

For a receiving antenna, antenna gain is the ratio ofthe power delivered by the antenna in response to asignal arriving from a given direction compared tothat delivered by a hypothetical isotropic referenceantenna. The spatial variation of an antenna’s gain isreferred to as the radiation pattern or the receivingpattern. Actually, under the antenna reciprocitytheorem, these patterns are identical for a givenantenna. The receiver operates best with only a smalldifference in power between the signals from thevarious satellites being tracked and ideally theantenna covers the entire hemisphere above it withno variation in gain. This has to do with potentialcross-correlation problems in the receiver and thesimple fact that excessive gain roll-off may causesignals from satellites at low elevation angles to dropbelow the noise floor of the receiver.

Space borne systems at L-Band typically use circularpolarization (CP) signals for transmitting and receiving,

satellites orbiting the Earth don’t cause polarizationfading as they do with linearly polarized signals andantennas. Furthermore, circular polarization does notsuffer from the effects of Faraday rotation caused by theionosphere. Faraday rotation results in an electromag-netic wave from space arriving at the Earth’s surfacewith a different polarization angle than it would have ifthe ionosphere was absent. This leads to signal fadingand potentially poor reception of linearly polarizedsignals. GNSS satellites use right-hand circular polari-zation (RHCP) and therefore a GNSS antenna receivingthe direct signals must also be designed for RHCP.Antennas are not perfect and a RHCP antenna will pickup some left-hand circular polarization (LHCP) energy,referred as the cross-polar component.

Axial ratio is the measure of the polarization ellipticityof an antenna designed to receive circularly polarizedsignals. For high-end GNSS antennas such as choke-ringand other geodetic-quality antennas, the typical AR alongthe bore sight should be not greater than about 1 dB. ARincreases towards lower elevation angles and you should

Chokering antenna and RS data

R2 = 0.1427

20

30

40

50

60

70

80

0 5 10 15 20 25 30 35 40

No. of observations (000 & 1200 UTC)

IPW

V (

mm

)

Choke

RS

Linear (Choke)

RMSE=4.36 mm

Fig. 3 IPWV (mm)processing results byChoke-ring antennawith RS data

Patch antenna and RS data

R2 = 0.1281

20

30

40

50

60

70

80

0 5 10 15 20 25 30 35 40

No. of observations (000 & 1200 UTC)

IPW

V (

mm

)

Patch

RS

Linear (Patch)

RMSE=5.02 mm

Fig. 4 IPWV (mm) pro-cessing results by Patchantenna with RS data

J Indian Soc Remote Sens (September 2012) 40(3):389–398 393

look for an AR of less than 3–6 dB at a 10° elevationangle for a high-performance antenna.

The phase center is the point in space where all therays appear to emanate from (or converge on) theantenna or it is the point where the electromagneticfields from all incident rays appear to add up in phase.Ideally, this phase center is a single point in space for alldirections at all frequencies. However, a “real-world”antenna will often possess multiple phase center points(for each lobe in the gain pattern, for example) or a phasecenter that appears “smeared out” as frequency and

viewing angle are varied. For well-designed high-endGNSS antennas, phase center variations in azimuth aresmall and on the order of a couple of millimeters. Thevertical phase offsets are typically 10 mm or less. High-performance low noise amplifier (LNA) between theantenna element itself and the receiver is required to pickup the signals from space is on the order of −130 dBm.The performance of a particular receiver element isevaluated by its noise figure (NF).

Choke-ring ground planes are typical in geodeticantennas. These allow good gain pattern control,

Fig. 5 L1 frequency SNRsky plot (AR25-3D Chokering antenna)

Fig. 6 L1 frequency SNRtime Series (AR25-3DChoke ring antenna)

394 J Indian Soc Remote Sens (September 2012) 40(3):389–398

excellent multi-path suppression, high front-to-backratio, and good AR at low elevation angles. Chokerings contribute to a stable phase center.

Cycle Slips

The loss of carrier-phase tracking resulting in aninteger number of cycle’s discontinuity will cause anerror in the carrier-phase measurements. The loss maybe due to internal receiver tracking problems or an

interruption in the ability of the antenna to receive thesatellite signals (Seeber 1993).

The residual phase errors due to GPS antennas willnot only affect the precision in GPS networks withdifferent types of antennas, but also in networks usingidentical antennas if the network covers a large spatialarea (baseline lengths ~1,000 km) (Schupler andClark 1991).The performance of GPS antennasdepends on the differences in the design of theantennas, manufacturing variability between antennas

Fig. 7 L2 frequency SNRsky plot(AR25-3D Chokering antenna)

Fig. 8 L2 frequency SNRtime Series (AR25-3DChoke ring antenna)

J Indian Soc Remote Sens (September 2012) 40(3):389–398 395

of the same model, the material which surrounds theantenna (including the antenna mount and radome),and the frequency range over which the antenna isused (Schupler and Clark 2001).

IPWV Estimation

The estimated IPWV in mm for both choke ring andpatch antenna is shown in Figs. 3 and 4. The RMSEvalues in the case of choke ring are reduced

considerably as compared to the patch antennabecause patch antenna is not susceptible to suppressthe multi-path effect.

Signal to Noise Ratio (SNR) Sky Plots and Its TimeSeries

The geometry of the satellite (i.e. how closely or farapart satellites are spaced across the sky) is therepresentation of satellite availability from receiver

Fig. 9 L1 frequency SNRsky plot: Trimble Zephyrantenna (TRM 39105.00)

Fig. 10 L1 frequency SNRtime Series: Trimble Zephyrantenna (TRM 39105.00)

396 J Indian Soc Remote Sens (September 2012) 40(3):389–398

point of view, which is graphically termed asskyplot. Sky plot is a plot of satellite tracks on azenithal projection centered at the GPS groundstation or it is panoramic view how the satellitesare distributed. The SNR plots of L1 and L2

frequencies for Leica 3D choke ring and TrimbleZephyr antenna are shown in Figs. 5, 7, 9 and 11.The average signal strength is represented with acolor bar along with its value. The numerical valuesare given in Tables 1, 2 and 3. The error contribu-

tions of available satellites for both frequencies areshown in Figs. 6, 8, 10 and 12.

Suitability of Choke Ring Antenna

Choke ring antenna configuration allows betterreception and improved multi-path rejection capabil-ity. This is because of the design of its ground planewhich composed of three dimensional arrays of threeto five concentric slots and helps in reduces the multi-

Fig. 11 L2 frequency SNRsky plot: Trimble Zephyrantenna (TRM 39105.00)

Fig. 12 L2 frequency SNRtime series: Trimble Zephyrantenna (TRM 39105.00)

J Indian Soc Remote Sens (September 2012) 40(3):389–398 397

path effect. The choke rings are usually a quarterwavelengths deep, which creates high impedancesurface to prevent propagation of surface waves nearthe antenna and excitations of undesired modes. Inthis way, the data gaps and cycle slips are reduced dueto very smooth, minimal ripple and phase centerstability of the antenna.

Conclusions

& The IPWV estimates are better from choke ringantenna having R2=0.1427 and RMSE=4.36 mmas compared to patch antenna R2=0.1281 andRMSE=5.82 mm.

& LeicaAR25 3D Choke ring antenna has bettersignal strength and suppression of multi-patheffects as compared to the LeicaX1202, andTrimble TRM 39105.00.

& Data gaps, cycle slips are lesser in LeicaAR25 3DChoke ring antenna as compared to LeicaX1202,and Trimble TRM 39105.00.

& The number of observations or Epoch is more incase of LeicaAR25 3D Choke ring antenna.

& LeicaAR25 3D Choke ring antenna supports themulti-constellation with optimum accuracy.

Acknowledgments The author is grateful to the DirectorGeneral of India Meteorological department, Lodi road, NewDelhi for providing the data to accomplish this task.

References

Estey, L. H., & Meertens, C. M. (1999). TEQC: The Multi-Purpose Toolkit for GPS/GLONASS Data, L GPS Solutions,vol 3 (pp. 42–49). Wiley.

Langley, R. B. (1998). A primer on GPS antennas. GPS World,9, 50–54.

Misra, P., & Enge, P. (2001). Global positioning system:Signals, measurements, and performance. Lincoln:Ganga-Januma Press. 390.

Moernaut, G. J. K., & Orban, D. (2009). Innovation: GNSSantennas. GPS World, 1, 23–29.

Schupler, B. R., & Clark, T. A. (1991). How different antennasaffect the GPS observable. GPS World, 2, 32–36.

Schupler, B. R., & Clark, T. A. (2001). Characterizing thebehavior of geodetic GPS antennas. GPS World, 12, 48–55.

Seeber, G. (1993). Satellite Geodesy: Foundations, methodsand applications. Berlin: Walter de Gruyter. 531.

398 J Indian Soc Remote Sens (September 2012) 40(3):389–398