Characterizing the Behavior of

Geodetic GPS Antennas

The performance of GPS user antennas isinfluenced by many factors. Some are

inherent to the design of the antenna, whileothers are “outside influences,” such as theeffect of material close to the antenna (includ-ing the antenna radome, if any), the designof any integral antenna amplifier (pream-plifier), and the frequency range over whichthe antenna is operated. The user must beaware of the impact that these factors will

48 GPS World February 2001 www.gpsworld.com

Characterizing the Behavior of

Geodetic GPS AntennasBruce R. Schupler Honeywell Technology Solutions Inc.

Thomas A. Clark NASA Goddard Space Flight Center

ing a variety of materials close to the anten-na, and performing the antenna charac-terization at frequencies that range frombelow the proposed third civil GPS fre-quency (L5 at 1176.45 MHz) to above thetop of the L1 GPS band. (This frequencyrange includes all of the GLONASS fre-quencies.) In this article we will describethe results of recent tests on three antennasof interest and available to us or, in one case,that a colleague asked us to test. (Antennamanufacturers answered technical ques-tions and received the test results, but wereotherwise not involved in designing thetests.)

Geodetic Antenna RequirementsOne of the many benefits of GPS is that itserves a large and varied community of users, from the weekend sailor who iscontent with positioning accuracies of a few tens of meters to the geodesist to whom even a few millimeters can make adifference.

The level of complexity of the hardwareand software needed by these two kinds ofGPS users is different. Whereas the week-end sailor may use a simple handheld GPS receiver providing positions based on single-frequency pseudorange mea-surements, the geodesist uses much moreelaborate equipment providing dual-frequency carrier-phase observations inaddition to low-noise pseudorange obser-vations. And whereas the weekend sailor’sreceiver may safely ignore small effectson its measurements, the geodesi st’s equipment and data processing softwaremust account for effects as small as a few millimeters. Phenomena such as the solidEarth tides and the effect of troposphericwater vapor on the propagation of the GPS signals, for example, must be modeledby the data processing software. The GPShardware itself also will have an effect onthe measurements. In addition to randomthermal noi se, the measurements will be affected by any instability in the phase center of the antenna and any multipath signals accepted by the antenna.

Two of the more important character-istics of antennas for high-accuracy GPSapplications such as geodesy are a highphase-center stability and a low responseto multipath signals. Several manufactur-ers have designed such antennas. How-ever, even these state-of-the-art antennasare not perfect, and it is useful to charac-terize their performance under different operating conditions. It is difficult to carryout such characterization tests in the field using “live” GPS signals, because

In high-accuracy applications of GPS such as establishing geodetic control networks, mon-itoring dam deformation, or measuring the Earth’s rotation, effects on GPS measurementsas small as a few millimeters can be important. To achieve the required posit ioningaccuracies, such ef fects must be modeled very carefully or preferably avoided in thefirst place. Although some potential errors originate with the GPS satellites and some withthe ionosphere and troposphere through which the signals must travel, some are due tothe receiver’s antenna and its immediate environment. High-accuracy applications use spe-cial antennas designed to reduce antenna-related errors to a minimum. Just how good arethese antennas? It is difficult to check the performance of antennas in the field — wherethe ground, mounting devices, and nearby structures all may have an ef fect. To isolatean antenna from its environment as much as possible or to change the environment in acontrolled fashion, antennas are tested in anechoic chambers — specially designed enclo-sures that virtually eliminate reflected signals and in which the position and orientationof the antenna can be precisely controlled. In this month’s column, Bruce Schupler andThomas Clark discuss the procedures they have developed to characterize the behaviorof GPS antennas using anechoic chamber measurements and discuss some of the resultsthey have obtained.

Bruce Schupler is the Program Manager for Satellite Laser Ranging and Very Long BaselineInterferometry (VLBI) at Honeywell Technology Solutions Inc., Lanham, Maryland. Hehas been involved in the development of space geodetic instrumentat ion and dataanalysis since 1977 with particular expertise in VLBI. He has been active in the high -accuracy characte r izat ion of GPS antennas for geodet ic applicat ions since 1989.Thomas Clark joined NASA’s Goddard Space Flight Center in Greenbelt, Maryland, in 1968.His research interests have centered on developing the techniques of VLBI and GPS for high-accuracy geodesy. His research also has involved the characterization of geodetic GPSantenna performance at millimeter levels and the mitigation of site-specific multipath. Clarkhas been author or co-author of more than 150 scientific and technical papers in manyfields. He has received numerous awards, including NASA’s Medal for Exceptional EngineeringAchievement.

Inno

vati

on

have on the performance of the antenna.For several years we have been charac-

terizing the performance of a variety of geo-detic-quality GPS user antennas using theanechoic chamber of the Goddard SpaceFlight Center (GSFC). We have recentlyexpanded our test program to address someof the outside influences on the perform-ance of the basic antenna. We have test-ed several antennas both with and withoutradomes, with and without amplifiers, plac-

www.gpsworld.com GPS World February 2001 49

the various factors affecting the observa-tions cannot be easily isolated. Instead,tests are conducted in a controlled envi-ronment — an anechoic chamber — usingsignals from a generator which mimics realGPS signals.

Measurement ProceduresPrevious versions of the measurement anddata analysis procedures that we followhave been discussed in some detail (seeFurther Reading). We will not repeat thisdetailed discussion here. However, a briefsummary of our current measurement processwill be useful for those who are unfamiliarwith our procedures.

We performed all of our measurementsin the anechoic chamber located in GSFCBuilding 19. This chamber was recentlyrebuilt and is now fully automated. It con-sists of a large antenna positioner locat-ed at the large end of a room built in theshape of a square horn with a source anten-na located 18 meters away at the throatof the horn. All of the interior surfaces ofthe anechoic chamber are lined with radio-frequency (RF) absorbent material.For our purposes, the source antenna is adipole that can be rotated under computercontrol to provide a signal that i s hori-zontally or vertically polarized. The anten-na positioner can rotate the antenna under measurement through 360 degreesof motion both in the plane of the anten-na (to provide azimuthal coverage) and ina plane perpendicular to the plane of theantenna under measurement (to provideelevation coverage). Figure 1 shows ourrange calibration standard GPS antennamounted on the antenna positioner in theanechoic chamber.

As the antenna under test is rotated over

the desired angular measurement range,the source signal is stepped through the var-ious frequencies and the amplitude andphase response of the antenna under testis measured. (For our recent tests, we used129 discrete frequencies.) This measure-ment process is repeated for horizontallyand vertically polarized source signals. Fromthis data, the measurement software cansynthesize the antenna response to rightand left circularly polarized signals as wellas the cross-polarization and axial ratioresponse functions. The amplitude responseis referenced to the response of a well-char-acterized standard gain horn.

To determine the position of the anten-na phase center relative to the base ofthe antenna, the offset between the pro-jection of the vertical axis of the antennapositioner and the front of the antennamounting fixture must be determined. Thisis established through the use of a laser, anauxiliary table, and a plumb bob as shownin Figure 2.

After collecting the magnitude and phasedata, sorting them, and extracting them intousable files , we have to process the phasedata to extract the antenna phase centerand to generate phase patterns that cor-respond to the computed phase center. Thisprocess consists of fitting the data to a modelthat takes into account the mechanical fea-tures of the antenna positioner as well asthe effects that the change in range betweenthe RF source and the antenna under testproduces as the latter is rotated. As endproducts of this data processing phase, weobtain files of antenna magnitude patterns,phase patterns, and phase center locationsfor each frequency that we measured atboth right and left circular polarization.

The recovered antenna phase center and,

FIGURE 1 The range calibra-tion antenna on the antennapositioner in the anechoicchamber

www.gpsworld.com GPS World February 2001 49

thus, the recovered phase pattern is a func-tion of the elevation angle cutoff that is usedto fit the phase center. For all of our work wehave used an elevation angle cutoff of 10degrees when fitting for the phase center.

Our two most recent antenna measure-ment sessions occurred in October 1998and November 1999. In Table 1 we list thegeodetic antennas and measurement con-figurations that we measured during thesesessions and that we discuss in this article.(In addition to the geodetic antennas, wealso measured several L1-only and otherexperimental antennas. We will not discussthose measurements here.)

Space constraints do not allow us to pre-sent all the results we produced from pro-cessing the data we obtained from the mea-surement sessions listed in Table 1. Instead,in what follows, we show selected resultsthat highlight the effects that various para-meters can have on the performance of geo-detic GPS user antennas.

Changes in Antenna Responsewith FrequencyThe choke ring style GPS antenna exhibitsvery significant changes in its responseas a function of frequency. Figure 3 showsthe amplitude response of several chokering antennas at the L1, L2, and L5 frequencies,and Figure 4 shows the change in phaseresponse.

As may be seen from those two figures,the shapes of the amplitude and phase pat-terns at the different frequencies for the fourdifferent antennas are generally similar. Thegain of the antennas varies from unit to unit(most likely due to differences in the inte-gral amplifiers) while all of the units showsubstantially lower gain at L5 than at L1 or L2.

FIGURE 2 The laser, auxil-iary table, and plumbbob used to determinethe antenna offset

FIGURE 3 The amplitude response of four choke ring antennasat L1, L2, and L5

–180 –90 0 90 180Zenith distance (degrees)

–20

–10

0

10

20

30

40

50

60

70

Am

plit

ude

(dB)

D&M TAshtechLeicaTrimble

L1 L2 L5

Innovation

50 GPS World February 2001 www.gpsworld.com50 GPS World February 2001 www.gpsworld.com

The phase patterns are also quite similarfor all of the antennas at a given frequency.The only exception to this is the slightlynoisy appearance of the Ashtech antennaat L5 around a zenith di stance of 650degrees.

How Similar Are Antennas fromDifferent Manufacturers?Although the amplitude and phase responseof choke ring style GPS antennas from var-

Effect of Reflectors and RadomesIn order to explore the effect that variousreflectors placed in the vicinity of a GPSuser’s antenna could have on the data col-lected by the antenna, we modified theantenna mounting structure shown in Figure 1 so that it could support a varietyof reflectors, as listed in Table 1. Figure 6shows the cementboard (cement filled panels) we used to simulate a concrete monument.

Figure 7 shows the effect that variousreflectors have on the L1 phase pattern ofthe Dorne & Margolin T antenna — a Dorne& Margolin antenna element on a JPL-designed choke ring with integral ampli-fier. The data plotted in Figure 7 is the changein measured phase between the describedconfiguration and the antenna with no addedreflector.

As can be seen in the figure, we obtainedvery similar and minimal effects with thecementboard and plywood reflectors, andwe obtained a quite large effect (as expect-ed) from the foil-faced insulation. We alsoexamined the effect of placing an absorberover the foil-faced insulation. It appears that the absorber did not completely shield

ious manufacturers are generally similar, thechange in the ver tical component of thephase center location with frequency doesvary somewhat — see Figure 5. The absolutevalue of the ver tical scale in the figure is arbitrary for each antenna and has beenadjusted independently for each antennafor plotting purposes. What is significant in thi s plot i s the differ ing shape of the curve for each antenna as a function of frequency.

FIGURE 4 The phase response of four choke ring GPS anten-nas at L1, L2, and L5

-80 -60 -40 -20 0 20 40 60 80Zenith distance (degrees)

-15

-10

-5

0

5

10

Phas

e (d

egre

es)

D&M TAshtechLeicaTrimble L1 L2 L5

TABLE 1 Geodetic GPS Antenna Configurations Recently Measured in the GSFCAnechoic ChamberAntenna ConfigurationAshtech Model 701945-01 With no added materialAshtech Model 701945-01 With 2-inch pipe adapterAshtech Model 701945-01 With short radomeAshtech Model 701945-01 With 2-inch pipe adapter and short radomeAshtech Model 701945-01 With 2-inch pipe adapter and radome bottomAshtech Model 701945-01 With 2-inch pipe adapter, radome bottom, and tall radomeDorne & Margolin T Amplifier removedDorne & Margolin T Foil placed over choke rings, amplifier removed — test of element onlyDorne & Margolin T S/N 198 Range standard antenna in normal configurationDorne & Margolin T S/N 198 Cementboard simulating a monument 11.5 cm behind antennaDorne & Margolin T S/N 198 30-cm-diameter reflecting disk mounted on cementboard behind antennaDorne & Margolin T S/N 198 Cementboard behind antenna and conical radomeDorne & Margolin T S/N 198 Cementboard behind antenna and hemispherical Goddard East GPS site

(GODE) radome. Dorne & Margolin T S/N 198 Foil-faced insulation as reflector 11.5 cm behind antennaDorne & Margolin T S/N 198 Absorber over foil-faced insulation 11.5 cm behind antennaDorne & Margolin T S/N 198 Plywood 11.5 cm behind antennaDorne & Margolin T S/N 198 Plywood behind antenna and conical radomeDorne & Margolin T S/N 198 Plywood behind antenna and hemispherical (GODE) radomeJPS RegAnt Dual Depth With radomeJPS RegAnt Dual Depth Without radomeJPS RegAnt Single Depth With radomeJPS RegAnt Single Depth Without radomeLeica AT504 No radomeLeica AT504 With Leica radome

FIGURE 5 Different antennas exhibit different vertical phasecenter motion as a function of frequency

1525.00 1551.75 1578.50 1605.25 1632.00Frequency (MHz)

0

1

2

3

4

5

6

7

8

Posi

tion (

mill

imet

ers

– ar

bit

rary

ori

gin

)

D&M T

AshtechLeicaTrimble

D&M T (no preamp)

FIGURE 6 Modified antenna mountshowing the cementboard used tosimulate a concrete monument

Inno

vati

on

www.gpsworld.com GPS World February 2001 51

In addition to performing this test on theAshtech antenna, we also tested a LeicaModel AT504 antenna with and withoutits radome. This antenna also showed a low-ering of its phase center by 2 to 3 millime-ters when the radome was installed. (Wewill discuss later the cause of the oscilla-tion in the phase center vertical position

www.gpsworld.com GPS World February 2001 51

the antenna from the effect of the foil. Thisis somewhat surprising, because we usedthe same absorber that we used on the wallsof the anechoic chamber.

The 12-inch (30.48 centimeters) metaldisk we placed in the middle of the cement-board simulated a metal top plate on a concrete pier. Although this did have a different effect on the antenna than thecementboard alone, the difference was small.

To explore the effect of radomes on theperformance of the GPS user antenna, weperformed our series of measurements usingan Ashtech Model 701945-01 antenna, mount-ing adapter, short radome, and tall radome.Although the phase and amplitude patternsdid not change noticeably as we added the various components to the basic antenna, the location of the phase centerdid vary. This variation in the frequencyrange around L1 is shown in Figure 8. (Theresults in the frequency range near L2 and L5 have a magnitude similar to thosenear L1.)

In Figure 8 we translated the traces withthe word “shifted” in the label by the heightof the 2-inch (5.08 centimeters) pipe adapterto make them fit onto the same scale as theother traces. The origin of the phase centerposition in thi s plot i s arbitrary and i s not tied to any physical feature of the antenna. The vertical scale is meant only toshow the effect that adding componentsto the basic antenna has on the variation ofthe phase center vertical component withfrequency.

As expected, the most significant changesin the vertical position occur when eitherradome is added. This effect appears to bea lowering of the phase center positionby approximately 2 millimeters.

seen at the high-frequency end of Figure 8.)

The Effect of a Change in DesignSo far we have discussed antennas of basi-cally the same design. The Javad PositioningSystems (JPS) antennas are of a somewhatdifferent design and exhibit somewhatdifferent characteristics than the “normal”

Innovation

FIGURE 7 The effect of reflectors on the phase pattern of theDorne & Margolin T antenna at L1

–80 –60 –40 –20 0 20 40 60 80Zenith distance (degrees)

–5

–3

–1

1

3

5

7

Phas

e (d

egre

es)

FoilboardAbsorber over foilboardPlywoodCementboardCementboard + 12-inch diskRemeasurement with no material added

FIGURE 8 The effect of radomes and antenna mounts on thephase center position of the Ashtech 701945-01 antenna

1525.00 1551.75 1578.50 1605.25 1632.00Frequency (MHz)

0

1

2

3

4

5

6

7

8

9

10

Pos

ition

(m

illim

eter

s)

Antenna shifted2-inch adapterShort radome shifted2-inch and short radome2-inch and radome bottom2-inch and tall radome

Circle 32

occurs within the L2 passband. When theradome is installed, this feature is reducedin magnitude and shifted out of the L2 pass-band. This clearly shows that the radomein this antenna is not an auxiliary item.Rather, it is part of the RF structure of theantenna. Interestingly, for the RegAnt SingleDepth antenna, Figure 10 shows that in theL1 area the addition of the radome makesthe phase center move up rather than down.This clearly indicates that the radome plays

52 GPS World February 2001 www.gpsworld.com

choke ring antennas. We tested three JPSantennas:a RegAnt Single Depth, a RegAntDual Depth, and a LegAnt. We will reporthere the test results for the two variantsof the RegAnt, both of which we tested withand without their integral radomes.

The phase patterns of all configurationsof the RegAnt are similar to those of ourrange standard Dorne & Margolin T anten-na at L2 and L5 and of noticeably smallermagnitude at L1. However, the most strik-

ing difference we saw between the perfor-mance of the RegAnt antennas and thatof our range standard antenna was whenwe plotted the phase center vertical posi-tion as a function of frequency. We showthis information in Figures 9 and 10.

The most obvious feature of Figure 9 isthe rapid change in the phase center posi-tion of the RegAnt Single Depth antenna inthe vicinity of 1200 MHz. When the radomeis not installed on this antenna, this change

FIGURE 9 The vertical phase center position of the JPS anten-nas as compared with the Dorne & Margolin T near L2 and L5

1166.00 1193.25 1220.50 1247.75 1275.00Frequency (MHz)

60

70

80

90

100

110

120

Pos

ition

(m

illim

eter

s)

RegAnt Dual no domeRegAnt Dual with dome

RegAnt Single no dome

D&M T (shifted)D&M T (no preamp, shifted)

RegAnt Single with dome

FIGURE 10 The vertical phase center position of the JPSantennas as compared to the Dorne & Margolin T near L1

1525.00 1551.75 1578.50 1605.25 1632.00

Frequency (MHz)

60.00

68.75

77.50

86.25

95.00

Pos

ition

(m

illim

eter

s)

RegAnt Dual no domeRegAnt Dual with dome

RegAnt Single no dome

D&M T (shifted)D&M T (no preamp, shifted)

RegAnt Single with dome

ANTENNA TERMINOLOGYAnechoic chamber. An enclosure ranging in size from a few meters totens of meters on a side used for the testing of antennas and other RFdevices. The interior walls of the chamber are covered with RF-absorb-ing material that reduces signal reflections or “echoes” to a minimum.

Axial ratio. A measure of the polarization ellipticity of an antennadesigned to receive circularly polarized signals. An axial ratio of unity,or 0 dB, implies a perfectly circularly polarized antenna.

Choke ring. An antenna ground-plane consisting of several concen-tric metal hoops, or thin-walled hollow cylinders, mounted on a circularbase at the center of which is placed an antenna element such as amicrostrip patch. It significantly attenuates ground-bounce and low-ele-vation angle multipath.

Cross-polarization response. A measure of the degree to whichan antenna designed for, say, right-hand circularly polarized signals,responds to left-hand circularly polarized signals.

Dipole. A simple directional antenna which consists of a linear con-ductor, often wire or thin metal tubing, approximately one half-wave-length long, with the feed point in the middle. It is often used as a refer-ence antenna.

Gain. For a transmitting antenna, the ratio of the radiation intensityin a given direction to the radiation that would be obtained if the poweraccepted by the antenna was radiated isotropically. For a receivingantenna, it is the ratio of the power delivered by the antenna inresponse to a signal arriving from a given direction compared to thatdelivered by a hypothetical isotropic reference antenna.

Gain (amplitude) pattern. The spatial variation of an antenna’sgain.

Horn antenna. An aperture antenna, typically in the shape of a trun-cated cone or pyramid, used at microwave frequencies. It is essentially awaveguide that is flared toward the open end from which electromag-netic waves are launched (or received, in the case of a receiving anten-na). The simple geometry of the horn antenna permits accurate calcula-tion of its gain pattern and is noted for its low level of backlobes.

Phase center. The apparent source of radiation of a transmittingantenna. If the source is an ideal point source, the phase center is thecenter of the radiating spherical wavefronts (of equal phase). For a GPSreceiving antenna, it is the point to which the receiver’s phase measure-ments actually refer. Since a real antenna is not an ideal point source,its equiphase contours will not be perfectly spherical, and hence thecenter of curvature may vary with the azimuth and elevation angle of anarriving signal.

Phase pattern. The spatial variation of an antenna’s phase center.Polarization. The sense of vibration of electromagnetic radiation.

There are two main types of polarization: linear, in which the radiatingwave’s electric field vector is confined to a particular direction (typicallyvertical or horizontal); and circular, where the electric vector rotates asthe wave propagates through space. Depending on the sense of rota-tion, a signal’s waves may be left-hand or, as with GPS signals, right-hand circularly polarized. For maximum response, the polarization of areceiving antenna must match the polarization of the signals.

Radome. An electromagnetically transparent cover intended to pro-tect an antenna from the effects of its physical environment.

RF absorber. Material, such as carbon-impregnated foamed plastic,that absorbs (rather than reflects) incident radio signals.

Inno

vati

on

54 GPS World February 2001 www.gpsworld.com

a significant role in the RF design of thisantenna.

The RegAnt Dual Depth antenna showsalmost no effect from the radome in the L2area and a small effect near L5. However,the radome has a significant impact on theantenna in the L1 region. Once again, theradome forms a portion of the antenna’s RFstructure. Neither version of the RegAntshould be operated without its radome.

What Limits the FrequencyResponse of the Antennas?We wished to address the question of theeffect of the integral amplifier on the responseof otherwise very similar GPS antennas.Based on measurements that we performedseveral years ago in preparation for usinga Dorne & Margolin T antenna for combinedGPS and GLONASS measurements, we sus-pected that the response of the amplifierwas limiting the bandpass of the antenna.We needed to explore this question furtherto determine whether or not we could usecurrent antennas at L5.

To determine the effect of the amplifi-er, we removed it from a Dorne & MargolinT antenna, ran the antenna through our measurement process, and compared the

ConclusionsWe presented here the data on the responseof several different geodetic GPS user anten-nas over a wide frequency range and inmany different configurations. Although wehave made no attempt to apply the anten-na calibrations that can be derived fromthis data to field-collected GPS data, theeffects shown here will doubtless also beseen in such situations.

Our measurements lead us to the fol-lowing conclusions:

c Similar antenna designs from differ-ent vendors perform in a generally similarmanner.

c Almost anything you put near anantenna affects its response.

c A change in an “auxiliary” portionof an antenna assembly (radome, amplifi-er, etc.) can significantly change theresponse.

c The performance of some antennadesigns depends critically on the cou-pling between the antenna and itsradome.

c The amplifier/bandpass filterresponse limits the L5 performance ofmany of the current choke ring antennadesigns.

results with those of our standard rangeantenna with an amplifier installed. Whenwe examined the results of this test, theeffect of the amplifier became quite clear.As we show in Figure 5, the rapid oscilla-tions in the phase center vertical positionresult from the response of the amplifierrather than being an inherent character-i s tic of the choke r ings or the antennaelement. Although the amplitude responseshows a very significant, albeit smooth,decrease at the extreme frequencies andthe phase patterns still appear to be well-behaved, the phase center in the case withthe amplifier is not well behaved. We sus-pect that this is caused by the bandpass filters within the amplifier being used at orbeyond their designed band edges at bothends of our frequency range. Indeed, thesame rapid change in phase center posi-tion can be seen in the frequency rangebetween L2 and L1. The Leica antenna usesa different amplifier design and does notexhibit this phase center vertical oscilla-tion to nearly as great a degree as the Dorne& Margolin T does, whereas the responseof the Ashtech is similar to that of the Dorne& Margolin T.

Inno

vati

on

This site is supported by...

Circle 26 Circle 27 Circle 28

www.gpsworld.comGet in Touch with theWorld of GPSMore than 32,000 people visit GPS World ’s Web site eachmonth

www.gpsworld.com GPS World February 2001 55

Additional results for the test configu-rations described in this article are avail-able on request.

AcknowledgmentsWe could not have performed this workwithout the assistance of Steve Seufert ofGSFC and his expertise with the GSFC ane-choic chamber and without the resource-fulness of Charles Kodak of Honeywell infinding just the right item to make the testspossible. In addition, we would like to thankthose manufacturers and organizations wholent us antennas to test (see next section).

This measurement program is fundedunder the Solid Earth and Natural HazardsNASA Research Announcement to improvethe inherent accuracy of GPS measurementsused in Earth science research. We eager-ly seek the cooperation of industry andacademia to sample new antenna, radome,and mounting technologies for the sciencecommunity.

This article is based on the paper “HighAccuracy Characterization of Geodetic GPSAntennas Using Anechoic Chamber andField Tests” presented at ION GPS 2000, the13th International Technical Meeting of the Satellite Division of The Institute of

FURTHER READINGFor an introduction to GPS antennas andtheir characteristics, see

c “A Primer on GPS Antennas,” by R.B.Langley in GPS World, Vol. 9, No. 7, July1998, pp. 50–54.

For discussions about the effects anten-nas have on GPS observations, see

c “How Different Antennas Affect the GPSObservable,” by B.R. Schupler and T.A. Clark,in GPS World, Vol. 2, No. 10, November1991, pp. 32–36.

c “Signal Characteristics of GPS UserAntennas,” by B.R. Schupler, R.L. Allshouse,and T.A. Clark, in Navigation, Vol. 41, No. 3,1994, pp. 277–295.

c “Characterizations of GPS UserAntennas: Reanalysis and New Results,” byB.R. Schupler, T.A. Clark, and R.L. Allshousein GPS Trends in Precise Terrestrial,Airborne, and Spaceborne Applications, theProceedings of the International Associationof Geodesy Symposium No. 115, pp.328–332, Springer-Verlag, Berlin, 1996.

The International Association of Geodesyestablished a Special Study Group to investi-gate antenna and site effects on GPS mea-surements. For its final report, see

c <http://maia.usno.navy.mil/gpst/refs/ssg1158.html>.

Navigation, Salt Lake City, Utah, September19–22, 2000. cManufacturersThe antennas discussed in this article aremanufactured by Dorne & Margolin (now EDOCorporation, New York, NY); Javad PositioningSystems (JPS), (now Topcon Positioning Systems, SanJose, CA); Ashtech Precision Products Division ofMagellan Corporation, Santa Clara, CA);and LeicaGeosystems Inc. (Torrance, CA).

“Innovation” is a regu-lar column featuringdiscussions aboutrecent advances inGPS technology andits applications aswell as the fundamen-tals of GPS position-ing. The column is

coordinated by Richard Langley of theDepartment of Geodesy and GeomaticsEngineering at the University of NewBrunswick, who appreciates receiving yourcomments as well as topic suggestions forfuture columns. To contact him, see the“Columnists”section on page 4 of thisissue.

Circle 33