Antennas

Receivers
Title	Antennas
Author(s)	GMV
Level	Advanced
Year of Publication	2011

From a user perspective, antennas are the main interface between the GNSS Space Segment (the satellite constellations) and the User Segment (GNSS receivers), as they are responsible for capturing the L-band signals transmitted from space. Antennas requirements can range from gain vs. azimuth and elevation, multipath and interference rejection, or stability of the electrical phase centre, to size, shape and environmental constraints.

Typically, the antennas used in GNSS receivers are formed by a radiating element, mounted on a ground plane, followed by an amplifier and protected by a radome. The radiating element is responsible for the antenna bandwidth and radiating characteristics, the ground plane conditions the radiation pattern shape, foremost at low elevation angles, the amplifier sets the receiver noise figure, and the radome affects the phase centre. Due to the characteristics of GNSS systems, receiver antennas are typically Right-Hand Circularly Polarized (RHCP) antennas, and the spacial reception pattern is near hemispherical: this configuration enables user reception of satellite signals in any azimuthal direction, and from zenith to horizon, thus maximizing satellite visibility.

Antenna configurations

Different applications may require different types of antenna design, technology and location. In fact, as show in Figure 1, antennas can range from large roof-mounted antennas (e.g. for a static open sky view) to embedded antennas in receiver boards (e.g. for mobile handheld devices).

The most common types in terms of implementation of antenna elements, ground planes, are:

Figure 1: Examples of GNSS antenna types: rooftop Radome (left), embedded Patch (top-right), Choke Ring design (bottom-right).

Figure 2: View of the antennas at Galileo Control Centre in Fucino, Italy (Contents ESA).

• Patch Antennas: Patch antennas are often used in handheld mobile devices, such as portable navigation devices or smartphones, since the antenna is built as a low-profile thin microstrip (i.e. patch), suitable for small form-factor devices. Microstrip patches or aperture-coupled patches can be single frequency or dual frequency, but also broadband. Patches are not circularly polarized by nature, so they achieve circular polarization by quadrature excitation of two linearly polarized ports. The quality of the circular polarization depends on a major way on the details of the feeding network and may be sensitive to fabrication issues, and the crosspolarization near the horizon tends to be high. They provide an excellent form factor (size and weight), and are used in most aviation GNSS antennas satisfying ARINC 743 standard.

• Turnstile Antennas: The turnstile antenna consists of two dipoles fed in 90° phase. A horizontal turnstile antenna in free space is vertically polarized in its plane, and has opposite circular polarizations above and below. Therefore, for geodetic applications, it requires a ground plane. When combined with a choke ring ground plane, it is a very common element in geodetic applications.

• Helical Antennas: Helical antennas are typically built as a conducting wire around a cylindrical or conical core, in a helix form, and can be used with or without a ground plane, depending on the configuration. In GNSS receivers, the helix antennas can be used as directional antennas, with a radiation pattern along the axis of the antenna.

• Spiral Antennas: The equiangular planar spiral is one of the prototypical "frequency-independent" antennas, so it offers excellent bandwidth. The antenna is circularly polarized by nature, although the quality of the polarization may depend on the feeding arrangement. On the other hand, it is comparatively large. Since the basic element radiates equally in both hemispheres, a ground plane of some kind is required.

• Radome Antennas: As the name suggests, in this configuration the antenna elements are housed in a hemispherical-like dome (the radome). Radomes are usually added to protect the antenna when the site where it is installed demands it (e.g. in roof-mounted installations) due to weather conditions, antenna security, or wildlife concerns. Typically, hemispherical radomes are used, because they have little effect on the phase centre stability and symmetry, and non-hemispherical radomes are usually avoided, except when the shape is required by site characteristics (e.g. conical radomes for snow rejection). Antenna phase centre calibration should be performed with the radome.

• Choke Ring Antennas: A choke ring antenna design consists in surrounding a central antenna element with several concentric conductive rings, enclosed in a protective dome. Its design is notable for the ability to reject multipath and low elevation signals (including reflections on the ground). Choke Ring antennas reflects signals that come from below, and draw surface wave signals into the choke channels, where they are cancelled or reflected away from the receiving element.

• Resistive Plane Antennas: The resistive plane antenna uses electrical resistance (rather than frequency-tuned rings as Choke Ring antennas) to keep unwanted signals from reaching the antenna element. Signals that strike the plane lose all energy before they reach the element and cause interference. Resistive Plane antennas provide similar performances as Choke Ring antennas, but with less weight and cost.

• Phased-Array Antennas: Phased-array antennas are mostly used for military or high-end applications, and consist of numerous antenna elements arranged in an array pattern. Each signal from an antenna in the array can be processed by introducing dynamic phase shifts, and when the outputs of the array are summed, the effective radiation pattern can be maximized in some directions and nulled in other (e.g. to insert nulls in the antenna pattern to adjust to jamming threats).

• Parabolic Antennas: Resembling popular TV and communications satellite dish antennas, parabolic antennas use a parabolic reflecting surface to direct the radio waves. The parabolic antenna is highly directive, so RF signals coming from the satellites can be received from one particular direction only. Its directive build structure and high gain make this type of antenna suitable for Ground Segment stations, such as monitoring or uplink stations, as shown in Figure 2.

Applications

An alternative distinction of GNSS antennas is in terms of the envisaged application:

• Geodetic Antennas: Survey and geodetic applications typically require fixed high-accuracy receivers and antennas, so it is common for multi-frequency / multi-constellation choke-ring antennas to be used for best performance.

• Rover Antennas: Rover antennas are used in land survey, forestry, construction, and other portable or mobile applications, and represent a good trade-off between accuracy and portability. Due to the need for mobility, the antenna is usually mounted on a handheld pole, stand or tripod: nevertheless, due to size and weight constraints, choke ring antennas are not a good solution for rover applications, so a high-end alternative must be found to ensure the necessary accuracy.

• Handheld Receiver Antennas: These antennas are found in most current smartphones and portable navigation devices, and are chosen for size, cost and power consumption constraints. Usually single-frequency antennas, they are available in a range of implementations such as helical or patch antennas, both passive and active (see section below for details).

Characteristics

Antennas used in GNSS receivers are targeted at capturing L-band signals originating from satellite constellations, so care must be taken when matching design specifications in terms of signal bandwidth, centre frequency, desired response, transfer function, multipath mitigation, and electromechanical factors. Table 1 shows a few relations between antenna applications and characteristics.

Table 1: GNSS antenna applications and characteristics.

	Geodetic	Rover	Handheld
Frequency Bands	Single to multiband	Single to multiband	Singleband
Bandwidth	Broadband	Narrow to Broadband	Narrowband
Radiation Pattern	Controlled	Controlled	Not controlled
Multipath suppression	High	Medium	None
Sensitivity	High	Medium to high	Low
Interference Handling	High rejection	Good rejection	Minimal rejection
Phase centre	Very important	Important	Not important
Dimensions	Large	Portable	Very small
Weight	Heavy	Portable	Lightweight
Cost	High	Medium	Low

Antenna Power

In general, GNSS antennas don't need any electrical power to operate, and are referred to as passive antennas. Typical scenarios for passive antennas are mobile devices, where power consumption is a critical issue. However, passive antennas must connect to the receiver's front end with very low losses, since there is no amplification mechanism involved at signal reception.

In applications in which the antenna can be away from the receiver (e.g. rooftop antennas), signal amplification can be implemented in the antenna itself, and for that fact they are called active antennas. Their purpose is to amplify the received signal to compensate for potential losses in long coaxial cables, at the cost of higher power consumption (power is usually delivered from the front end through the coaxial cable itself). In this approach, a Low-Noise Amplifier (LNA) is usually added to the antenna setup, in order to ensure a low noise figure in the signals to process. An example of active antenna use is for indoor reradiating purposes, where an active antenna inside a building reradiates the signals coming from another antenna, placed outside the building.

Radiation Pattern

Figure 3: Example of a patch antenna pattern.

GNSS signals use Right Hand Circular Polarization (RHCP), but antennas are not perfect, so a RHCP antenna will pick up some Left Hand Circular Polarization (LHCP) energy as well. In radiation terms, RHCP is referred to as the co-polar component, and LHCP is referred as the cross-polar component. The typical radiation pattern coverage is around 160º, resulting in low gain for signals coming in at low elevation angles, as shown in Figure 3 for a patch antenna. This pattern aims at maximizing the gain at zenith, decreasing to near unity at an elevation angle of around 10º-15º, and to negative gains below the elevation threshold (sometimes referred to as elevation mask).

Real-world GNSS antennas experience a gain roll-off of 10 to 20 dB from broadside (90º elevation) to the horizon (0º elevation). At zenith, the ratio of the vertical electric field to the horizontal electric field response is near unity, and this ratio is referred to as the axial ratio (AR), defined as:

[math]\displaystyle{ AR(\theta) = 20.log \left ( \frac{E_{RHCP}(\theta) + E_{LHCP}(\theta)}{E_{RHCP}(\theta) - E_{LHCP}(\theta)} \right ) }[/math]

The axial ratio (AR) increases towards lower elevations, and is less than 3 to 6 dB at 10° elevation for a high-performance antenna.

Multipath

Signals sent from GNSS satellites will directly arrive at the antenna but they also may be reflected off the ground, buildings or other obstacles, arriving at the antenna multiple times and delayed in time. This phenomenon is called multipath, and since it degrades position accuracy, it should be avoided or reduced. Multipath signals can basically come from three directions: the ground, hitting the back of the antenna, the ground or an object, hitting the antenna at a low elevation, or an object, hitting the antenna at a high elevation.

In specular reflections, the reflected signal changes its polarization with respect to the original one: in other words, a RHCP signal will be reflected as a LHCP signal. The multipath susceptibility of an antenna can then be quantified, with respect to the antenna’s radiation pattern characteristics, by the multipath ratio (MPR). The multipath ratio is computed as the ratio between the RHCP gain at the satellite signal impinging angle [math]\displaystyle{ \theta\, }[/math] with respect to the sum of the RHCP and LHCP gains at the multipath signal impinging angle (180º - [math]\displaystyle{ \theta\, }[/math] for ground reflections; [math]\displaystyle{ \theta\, }[/math] for vertical object reflections):

[math]\displaystyle{ MPR_{ground} = \frac{E_{RHCP}(\theta)}{E_{RHCP}(180^\circ - \theta) + E_{LHCP}(180^\circ - \theta)} }[/math]

[math]\displaystyle{ MPR_{vertical} = \frac{E_{RHCP}(\theta)}{E_{RHCP}(\theta) + E_{LHCP}(\theta)} }[/math]

Gain roll-off towards the lower elevations helps to suppress multipath signals: however, a good AR is always a must, since gain roll-off alone is insufficient.

Phase Centre

Another important design factor for antennas is the phase centre. In fact, when a receiver determines a position solution, it is actually estimating the position of the electrical phase centre]] of the antenna, where the signals are captured, regardless of the distance from the antenna the receiver device really is. This electrical phase centre should not be mistaken with the physical phase centre: although the latter can be observed and is a physical constant, the first may vary, depending on the direction of arrival of the received signal.

Determining the phase centre is important in GNSS systems, particularly when millimeter positioning resolution is desired; nevertheless, for well-designed high-end GNSS antennas, phase centre variations in azimuth are small and in the order of few milimeters (the vertical phase offsets are typically 10 mm or less). See the articles on antenna phase centre and receiver antenna phase centre for more details.

Interference Handling

GNSS receivers are good at mitigating interference, but it is also essential to keep unwanted signals out of receivers as much as possible. An out-of-band interferer is an RF source outside of the GNSS frequency bands, which can drive the LNA into its nonlinear range, and the LNA starts to operate as a multiplier or comb generator. RF filters can reduce the out-of-band signals by tens of dBs, which is enough in most cases. Of course, filters add insertion loss, amplitude and phase ripple, all of which degrade receiver performance. In-band interferers can be the third-order mixing products, or simply an RF source that transmits inside the GNSS bands. If these are relatively weak, the receiver will handle them; from a certain power level on, however, commercial receivers can do nothing about them.

Noise and Out of Band Rejection

The signals received by the antenna are very weak. For instance, in GPS, the minimum received power is –158.5 dBW (for L1) and –164.5 dBW (forL2), for signals arriving at 5º elevation. Furthermore, the cable and the receiver elements generate noise that can affect the signal. To avoid this, the first element following the antenna should be a Low Noise Amplifier (LNA), so that the power of the GNSS signal is greater than the noise generated by the cable and the receiver elements.

Adding the fact that Out of Band signals need to be rejected or attenuated to prevent the saturation of the amplifiers, and that the antenna does not provide a good close Out of Band rejection, a pass band filter can be placed before the first amplifier. This operation, however, will degrade the Noise Figure, so there is a trade-off in the design of the antenna LNA between the Noise Figure and the Out of Band Rejection. Typically, the compromise between is reached by distributing the filtering across the LNA in two stages.

Signal Bandwidth

Besides electrical considerations, GNSS antennas must also take into account the signal structure in terms of RF spectrum and bandwidth, and should be selected not only in accordance with the front end specifications, but also with the envisaged usage and application. New constellations and modulations bring different spectrum allocations and bandwidths (see each constellation's articles for specifications on spectrum allocation and bandwidth). As an example, Table 2 illustrates GPS L1 and Galileo E1 bands (which share the same centre frequency), and shows several parameters related to the bandwidths of incoming signals.

Table 2: GPS and Galileo L1/E1 civil signal bandwidth characteristics.

	GPS L1 C/A (BPSK)	Galileo E1 B/C (BOC)	Galileo E1 B/C (CBOC)
Chip Rate [Mcps]	1.023	1.023	1.023 / 6.138
Primary Code Length [chips]	1023	4092	4092
Primary Code length [ms]	1	4	4
Symbol Rate [sps]	50	250	250
Centre Frequency [MHz]	1575.42
Receiver Reference Bandwidth [MHz]	4.092	8.184	24.552

For the GPS case, if a receiver only tracks C/A code on L1, the antennas (and receiver) need to accommodate bandwidths of approximately 4.092 MHz for a near-optimal bandwidth usage. For Galileo E1 signals, due to signal design and spread spectrum properties, the bandwidth doubles. Of course, the more frequency content of the received satellite signals is processed, the better the accuracy performance will be, so a dual-constellation receiver for GPS L1 and Galileo E1 would need only one antenna with the maximum bandwidth required.

Dual-frequency receivers, that enable determination of the ionospheric delay and provide robustness to interference, will need a dual-band antenna. Antenna bandwidth requirements and tradeoffs require solid knowledge of the application envisaged, the accuracy needed, and the available technology. For example, a survey receiver will need larger bandwidths to achieve accurate results, whereas a low-cost receiver may only have a antenna/front end bandwidth of around 2 MHz, but still receive enough signal power to determine position within meters. Although the receiver’s antenna/front-end bandwidth is directly proportional to the accuracy required for the application, it is also directly proportional to the processing load and power consumption for portable solutions.

Front end connection

Table 3: Example of coaxial cable types and losses [dB / 100m].

	915 MHz	1.2 GHz	2.4 GHz
RG-58	54.1	69.2	105.6
RG-8X	42.0	52.8	75.8
LMR-400	12.8	15.7	22.3
LMR-600	8.2	10.2	14.4
1 5/8" LDF	2.5	3.1	4.6

The antenna signal provided to the front end is received as a low-power signal, and is typically amplified and filtered before entering the heterodyning and sampling process. In the case of the antenna being far from the receiver, for physical constraints or application necessity, special attention should be made with the connecting cable.

Coaxial cables are built in many different ways, with different materials, with or without special shielding, and specify losses in units of power and length per frequency range - typical values are in the form of "X dB/100 m @ Y MHz". Requirements like cable length, impedance, antenna gain and power supply (for the case of active antennas) need to be taken into account to minimize signal losses. For illustrative purposes, an example of the loss ranges at stake is shown in Table 3. In this example, from a signal loss point of view, the "1 5/8" LDF" cable would have less losses in the 1.2 - 1.6 GHz range of GNSS signals, and should be considered for connecting an antenna to the front end.

Related articles

• Generic Receiver Description
• Receiver Types
• System Design Details