GEOG 862
GPS and GNSS for Geospatial Professionals

The Antenna

four images of different antennas, see text below
Top left: Microstrip patch antenna, Top right: Quadrifilar antenna
Bottom left: Dipole antenna, Bottom right: Helix antenna

Most receivers have an antenna built in, but many can accommodate a separate tripod-mounted or range pole-mounted antenna as well. These separate antennas with their connecting coaxial cables in standard lengths are usually available from the receiver manufacturer. The cables are an important detail. The longer the cable, the more of the GPS signal is lost traveling through it. These connecting coaxial cables are usually at standard lengths in an effort to make sure that the impedance of the trip through the cable can be calibrated to the receiver.

As mentioned earlier, the wavelengths of the GPS carriers are 19 cm (L1), 24 cm (L2) and 25 cm (L5), and antennas that are a quarter or half wavelength tend to be the most practical and efficient, so GPS antenna elements can be as small as 4 or 5 cm. Most of the receiver manufacturers use a microstrip antenna. These are also known as patch antennas. The microstrip may have a patch for each frequency. Microstrip antennas are durable, compact, have a simple construction and a low profile. The next most commonly used antenna is known as a dipole. A dipole antenna has a stable phase center and simple construction, but needs a good ground plane. A ground plane also facilitates the use of a microstrip antenna where it not only ameliorates multipath, but also tends to increase the antenna’s zenith gain. A quadrifilar antenna is a single frequency antenna that has two orthogonal bifilar helical loops on a common axis. Quadrifilar antennas perform better than a microstrip on crafts that pitch and roll, like boats and airplanes. They are also used in many recreational handheld GPS receivers. Such antennas have a good gain pattern, do not require a ground plane, but are not azimuthally symmetric. The least common design is the helix antenna. A helix is a dual frequency antenna. It has a good gain pattern, but a high profile.

In the illustration, there are four antenna styles or design types. In the upper left hand corner, the micro strip patch antenna – which is the most common. They can be as small as four or five centimeters, given the fact that the GPS carrier's wavelengths are right around 20 centimeters. The micro strip can receive one or all of the GPS carriers.

The next most common is the dipole, which is below. These pictures indicate just the type of antenna, not necessarily specifically GPS antennas. The dipole, you see there in the lower left. You may recall that this is the kind of antenna that was used with the Macrometer, the first commercial GPS receiver. The Dipole has a stable phase center, simple to construct. It needs a ground plane. You see microstrips and the other types with ground planes as well. A ground plane ameliorates multipath and it tends to increase the antenna's zenith gain-- in other words, the gain of the antenna straight up.

A quadrifilar antenna, which is in the upper right, is a single frequency antenna. It has two orthogonal, bifilar, helical loops on a common axis. They perform better on boats and airplanes, and things that pitch and roll and yaw. There are also many recreational handheld GPS receivers that use that type. Such antennas have a good game pattern and don't need a ground plane.

The least common is in the lower right. This is the helix antenna. It has a good game pattern, but a high profile.


Diagrams showing the power spectral density for L1 AND L2
L2 on Block II and L2 on Block IIR-M
Power Spectral Density (PSD) Diagrams for L1 and L2
Source: GPS for Land Surveyors

An antenna ought to have a bandwidth commensurate with its application. In general, the larger the bandwidth the better the performance; however, there is downside. Increased bandwidth degrades the signal to noise ratio by including more interference. GPS microstrip antennas usually operate in a range from about 2 to 20 MHz, which corresponds with the null-to-null bandwidth of both new and legacy GPS signals. For example, L2C and the central lobe of the C/A code span 2.046 MHz, whereas L5 and the P(Y) code have a bandwidth of 20.46 MHz. Therefore, the antenna and front-end of a receiver designed to collect the P(Y) code on both L1 and L2 would have a bandwidth of 20.46 MHz, but a system tracking the C/A code or the future L2C or the C/A code may have a narrower bandwidth. It would need 2.046 MHz for the central lobe of the C/A code, or if it were designed to track the future L1C signal, its bandwidth would need to be 4.092 MHz. A dual frequency microstrip antenna would likely operate in a bandwidth from 10 to 20 MHz.

These PSD diagrams illustrate the C/A and P code signals power per bandwidth in Watts per Hertz as a function of frequency. A microstrip antenna, the most common, usually operates in the range from about 2 to 20 megahertz, which, as you can see in the diagram, corresponds to the null-to-null bandwidth of both these existing, legacy GPS signals.

The L2C signal, like the C/A, has the span of 2.046 megahertz. You can see it in the highest portion or the central lobe of the diagram on the left. L5, like the P code, has a bandwidth of 20.46 megahertz, as you see in red in both illustrations. An antenna on the front end of the receiver has to be able to accommodate that bandwidth of 20.46 megahertz if it is to track all of these signals. If the system's tracking C/A code or the L2C only, it could have a narrower bandwidth.

Nearly Hemispheric Coverage

Diagram showing the GPS antenna pattern
GPS Antenna Pattern

Since a GPS antenna is designed to be omnidirectional, its gain pattern, that is the change in gain over a range of azimuths and elevations, ought to be nearly a full hemisphere, but not perfectly hemispheric. For example, most surveying applications filter the signals from very low elevations to reduce the effects of multipath and atmospheric delays. In other words, a portion of the GPS signal may come into the antenna from below the mask angle; therefore, the antenna’s gain pattern is specifically designed to reject such signals. Second, the contours of equal phase around the antenna’s electronic center, that is, the phase center, are not themselves perfectly spherical.

The gain, or gain pattern, describes the success of a GPS antenna in collecting more energy from above the mask angle, and less from below the mask angle. A gain of about 3 to 5 decibels (dB) is typical for a GPS antenna. Decibels do not indicate the power of the antenna, because the unit is dimensionless. It refers to a comparison. In this case, the gain of a real GPS antenna is measured by comparing it to a theoretical isotropic antenna. An isotropic antenna is a hypothetical, lossless antenna that has perfectly equal capabilities in all directions. Since a decibel is a unit for the logarithmic measure of the relative power, a 3 decibel increase indicates a doubling of signal strength, and a 3 decibel decrease indicates a halving of signal strength. This means that a typical omnidirectional GPS antenna with a gain of about 3 dB (decibels) has about 50% of the capability of an isotropic antenna. It is important that the GPS receiver antennas and pre-amplifiers be as efficient as possible, because the power received from the GPS satellites is low.  The minimum power received from the C/A code on L1 is about -160dBW, -160 decibel Watts and the minimum power received from the P code on L2 is even less at -166 dBW.

You can think of this diagram as a vertical slice through the gain pattern of the antenna. The gain pattern is really 3D. GPS antennas, of course, are intended to be omnidirectional but not perfectly hemispheric. For most applications, the GPS antenna wants to filter out the low elevation signals. And you could see how that works here, that it would not receive those that are at the lowest. This doesn't obviate the need for a mask angle, of course.

Just a brief description of a decibel-- we'll be seeing a little bit more of this again in some of the PSD diagrams. The decibel is a unit for logarithmic measure of relative power of a signal. It's a tenth of a bell, which was named for Alexander Graham Bell. dBW, decibel watt, indicates the actual power of a signal compared to reference of one watt. The decibel alone is dimensionless. It's a ratio.

Antenna Orientation

Diagram showing an antenna oriented North
Antenna Oriented to North
Source: UNAVCO

In a perfect GPS antenna, the phase center of the gain pattern would be exactly coincident with its actual, physical, center. If such a thing were possible, the centering of the antenna over a point on the earth would ensure its electronic centering as well. But that absolute certainty remains elusive for several reasons.

It is important to remember that the position at each end of a GPS baseline is the position of the phase center of the antenna at each end, not their physical centers, and the phase center is not an immovable point. The location of the phase center actually changes slightly with the satellite’s signal. For example, it is different for the L2 than for L1 or L5. In addition, as the azimuth, intensity, and elevation of the received signal changes, so does the difference between the phase center and the physical center. Small azimuthal effects can also be brought on by the local environment around the antenna. But most phase center variation is attributable to changes in satellite elevation. In the end, the physical center and the phase center of an antenna may be as much as a couple of centimeters from one another. On the other hand, with today’s patch antennas, it can be as little as a few millimeters.

It is also fortunate that the errors are systematic, and to compensate for some of this offset error, most receiver manufacturers recommend users take care when making simultaneous observations on a network of points, that their antennas are all oriented in the same direction. Several manufacturers even provide reference marks on the antenna, so that each one may be rotated to the same azimuth, usually north, to maintain the same relative position between their physical and phase, electronic, centers. Another approach to the problem’s reduction is adjusting the phase center offset out of the solution in postprocessing.

When performing a GPS control survey, the surveyor may notice an arrow on the ground plane of the antenna. This arrow is a guide to align a particular antenna with all of the antennas that are used in simultaneous observations of baselines. It makes it possible to orient all the antenna in the same direction. Generally, the antennas, let's say the antennas of three or four or receivers, are oriented to north. This alignment is important because the physical center of an antenna is not exactly coincident with the phase center of the antenna, and it's the phase center that is positioned by the GPS observation. That's where the signal is actually pinpointed. However, when we center an antenna over a point, we can only center it by its physical center. If the two were exactly coincident, it wouldn't be any concern, whatsoever, how the several antennas in an observation session are oriented. Since there is a difference between the two, the offset between the phase center and the physical center is always in the same direction at each point. Therefore, the baselines come out exactly as if the physical and the phase center were coincident.

For the more rough work, this sort of orientation is not vital. It's only when we're talking about control work, work that needs to be quite accurate, that all of the antennas need to be oriented to the same direction.

Height of Instrument

Diagram showing the height of an antenna when mounted on a tripod
Measuring the Antenna Height
Source: NOAA

The antenna's configuration also affects another measurement critical to successful GPS surveying - the height of the instrument. The measurement of the height of the instrument in a GPS survey is normally made to some reference mark on the antenna. However, it sometimes must include an added correction to bring the total vertical distance to the antenna’s phase center.

One of the errors that are quite avoidable but often corrupts the baseline measurements of a static GPS control survey is the antenna height-- the height of the instrument.

Here in this diagram, you see very many ways of measuring that height. It can be measured at slant height or measured with a tape, usually to the antenna reference point. The ARP, or the antenna reference point, is frequently the bottom of the mount of the antenna.

Of course, then, there's usually a correction that is needed to be added to actually bring that measurement up to the phase center of the antenna. This also becomes part of the necessary information when one is using continuously operating reference stations. We talked earlier about the fact that it is possible to take the downloads from NGS managed Continuously Operating Reference Station (CORS) stations that are available on the Internet and post-process the observations taken with a roving GPS receiver. If you are doing so, it is also necessary to know the height of the antenna at the CORS. Of course, that is not an antenna that you would've set up, but that information is available along with the files from the base station.