1. Introduction

Multipath is an important local error source in Global Navigation Satellite System (GNSS) measurements. It occurs when signals reach an antenna through both a direct path and one or more reflected or diffracted paths. Unlike many satellite orbit, satellite clock and atmospheric errors—which may be partly mitigated through differencing, estimation or correction models—multipath is strongly dependent on the antenna’s immediate surroundings and can differ substantially between installations.

Antenna design and installation are the first lines of defence. Antenna gain patterns and polarisation characteristics can attenuate some reflected signals before they enter the receiver. Receiver tracking and estimation algorithms can provide additional mitigation, but neither antenna design nor receiver processing can guarantee the removal of all multipath.

The appropriate antenna therefore depends on the application, required accuracy, signal bands, operating environment, installation constraints and calibration requirements. These factors are particularly important for RTK, surveying, geodetic monitoring and precision timing.

2. What is multipath?

Multipath occurs when a receiver processes a GNSS signal that contains the direct line-of-sight component together with delayed copies reflected from surfaces such as buildings, the ground, water, vehicles, solar panels or parts of the platform carrying the antenna. The additional path length introduces a delay; the reflected component also has a relative phase and amplitude that depend on the geometry and reflecting surface.

The receiver processes the composite signal. The reflected component can distort code-tracking correlation functions and perturb carrier-phase tracking, producing measurement errors. The magnitude and time variation of those errors depend on the signal waveform, excess path delay, relative amplitude and phase, antenna response, receiver bandwidth and tracking-loop or correlator design.

2.1 Effects on code and carrier-phase measurements

• Code pseudorange: Multipath can shift the estimated code-tracking point. Errors range from small values to metres or more in adverse conditions; no single range is a universal upper bound for all signals, geometries and receiver designs.
• Carrier phase: Multipath perturbs the measured carrier phase and can produce errors at the millimetre-to-centimetre scale in many precision applications. For a simplified single-reflection model, the phase error is commonly described as having a theoretical maximum of approximately one quarter of a carrier wavelength. This is about 4.8 cm at GPS L1 and 6.1 cm at GPS L2 when expressed as range-equivalent phase error. Real behaviour depends on the reflected signal and tracking conditions.

The following table gives qualitative guidance, not guaranteed performance limits.

Measurement or signal	General behaviour	Important qualification
Code on GPS L1 C/A	Can be strongly affected by multipath; errors may range from small values to metres in difficult environments.	Actual error depends on excess delay, reflection strength, waveform and receiver correlator.
Code on wider-band signals such as GPS L5 and Galileo E5	Generally offers better intrinsic code-multipath performance for comparable conditions.	Performance is signal- and receiver-dependent; wider bandwidth does not eliminate multipath.
Carrier phase	Often supports very precise measurements, but multipath can still cause millimetre-to-centimetre errors and affect ambiguity resolution.	The quarter-wavelength figure is a simplified theoretical bound, not a guaranteed bound on the final position solution.

GPS L5 has a chipping rate of 10.23 Mchip/s, corresponding to a chip duration of approximately 97.8 ns and a propagation distance of about 29.3 m per chip. Galileo E5 uses wideband AltBOC(15,10) modulation. These signal properties support sharper correlation features and generally improve code-multipath resistance, particularly for short-delay reflections. The receiver’s actual tracking architecture remains important.

2.2 Factors that influence multipath severity

• Reflector geometry and excess path delay: Nearby surfaces can create short-delay reflections that are difficult to distinguish from the direct signal. Longer-delay reflections may be easier for some specialised receiver algorithms to identify, but this is not guaranteed.
• Satellite elevation and azimuth: Low-elevation signals often encounter more local obstructions and reflecting surfaces, although multipath can affect satellites at higher elevations too. Elevation masking can reduce some errors but also reduces satellite availability and may worsen geometry.
• Signal waveform and receiver tracking: Signal bandwidth, modulation, correlator spacing, tracking-loop design and multipath-mitigation algorithms all influence the resulting measurement error.
• Reflection properties and polarisation: GNSS satellite signals are transmitted with right-hand circular polarisation (RHCP). A simple specular reflection can reverse circular-polarisation sense, but real surfaces and geometries may produce mixed polarisation. Antenna discrimination between RHCP and LHCP can help reject some reflected energy, but the benefit depends on angle, frequency and antenna design.
• Platform and installation: Roofs, vehicle bodies, masts, radomes, cables and nearby equipment can alter the antenna pattern or create additional reflections. The installation can therefore behave differently from the antenna measured in isolation.

3. Direct and reflected signal paths

Figure 1. The antenna receives the direct signal together with delayed reflected components. Their amplitudes, delays, phases and polarisation states depend on the reflecting surfaces and geometry.

4. How antenna design can reduce multipath

An antenna does not normally identify a reflection by its delay. Instead, its spatial response and polarisation characteristics can attenuate some signals according to their arrival direction and polarisation. Receiver processing may then provide further mitigation.

• Gain pattern and low-elevation response: Antennas designed for surveying often control gain near and below the horizon to reduce sensitivity to some ground and nearby-structure reflections. The trade-off is reduced reception of low-elevation satellites. The actual elevation-dependent gain pattern should be examined rather than inferred from the antenna category.
• Polarisation purity and axial ratio: Good RHCP performance and appropriate RHCP-to-LHCP discrimination can reduce the impact of some reflections. Axial ratio describes how closely the polarisation approximates circular polarisation; it is not, by itself, a complete measure of multipath rejection. Performance varies with frequency, azimuth and elevation.
• Ground plane: A suitable conductive ground plane can modify the radiation pattern and reduce reception from below, depending on its dimensions, geometry and mounting. Its edges can also cause diffraction, so the result is design-dependent.
• Choke rings: Concentric conductive grooves around a geodetic antenna element are designed to suppress surface currents and reduce low-elevation and ground-reflected signals. Choke-ring antennas are widely used at high-precision reference stations because of their stable, well-characterised behaviour and strong multipath mitigation. They are larger, heavier and generally more expensive than compact antennas; they are not automatically optimal for every installation.
• Phase-centre behaviour: The antenna’s effective phase centre varies with signal direction and frequency. Phase-centre offset (PCO) is the mean offset from the antenna reference point; phase-centre variation (PCV) describes changes around that model. For high-precision work, use suitable calibration data and apply the correct antenna model consistently. Calibration corrects antenna phase-centre behaviour; it does not remove arbitrary site multipath.
• Controlled-reception-pattern antenna (CRPA): A multi-element array can adapt its reception pattern to suppress signals arriving from selected directions. CRPAs are primarily used for interference and jamming mitigation. They may affect multipath in some circumstances, but they should not be treated as a general substitute for multipath-resistant antenna design and good siting.

Antenna calibration performed in a chamber characterises the antenna under controlled conditions. Nearby structures and the installed mounting configuration can still change the effective pattern and create site-specific multipath. For demanding applications, the complete antenna, mount, radome and installation should be considered.

5. Antenna types: practical comparison

Antenna type	Typical strengths	Typical applications and caveats
Compact patch (passive or active)	Small, low-cost and easy to integrate; performance varies widely by design and ground plane.	Consumer devices, asset tracking and embedded products. Do not assume precision multipath performance from size or band count alone.
Multi-band patch or helical	Can support multiple constellations and bands in a compact package; performance depends strongly on design and installation.	Drones, vehicles, robotics and compact precision systems. Verify phase-centre and low-elevation performance if accuracy demands it.
Survey/precision antenna with ground plane	Designed for more controlled gain and phase-centre behaviour than many compact antennas.	RTK rovers, machine control and surveying. Check calibration, mounting requirements and performance in the intended environment.
Choke-ring geodetic antenna	Strong multipath mitigation and well-established use in high-precision reference installations.	CORS and geodetic monitoring. Larger, heavier and more costly; select a calibrated model appropriate to the signals tracked.
Timing antenna	May be optimised for stable reception, suitable gain and fixed-site operation.	Timing receivers and synchronisation installations. Timing performance also depends on receiver design, sky visibility, antenna delay stability and installation.
CRPA / adaptive antenna array	Can form spatial nulls towards selected interference sources.	Jamming and interference resilience. More complex and not primarily a general-purpose multipath solution.

These categories are only a first screening tool. Compare measured gain patterns, polarisation performance, calibration data, supported frequencies, environmental ratings and test results for the specific antenna model.

6. Selection and installation

6.1 What to check in the datasheet

• Frequency and constellation coverage: Confirm support for every signal the receiver is intended to use, including the required L-band signals and any relevant E6 or BeiDou B3 signals. Check the exact frequency bands and polarisation requirements, not just a generic ‘multi-GNSS’ label.
• Radiation pattern: Review gain versus elevation and, where available, azimuth-dependent plots. Check low-elevation behaviour and performance below the horizon where relevant.
• Polarisation: Review axial-ratio and polarisation-discrimination data across the required bands and angular range. A single zenith value is not enough to characterise performance.
• Calibration: For centimetre-level or geodetic work, confirm that suitable absolute antenna calibration data are available in the format and convention required by the processing workflow. Ensure the antenna model and processing corrections are consistent.
• Reference point: Understand how the manufacturer defines the antenna reference point and phase-centre offsets. Apply the correct offsets and variations in the receiver or post-processing software.
• LNA, noise figure and filtering: For active antennas, confirm that gain is appropriate for cable loss and receiver input requirements. Consider noise figure, current consumption, out-of-band filtering and susceptibility to nearby interference. Excessive gain can overload a receiver front end.
• Mechanical and environmental suitability: Check mounting, connector type, cable routing, weatherproofing, vibration, temperature range and radome requirements. Confirm whether the published performance assumes a specific ground plane or mounting arrangement.

6.2 Installation practice

• Sky visibility: Choose a location with a clear sky view and minimal obstructions. Avoid nearby walls, parapets, metal roof edges, fences, solar panels and other reflective surfaces where practical.
• Reflector clearance: Keep the antenna away from large reflective surfaces and platform edges where the application permits. There is no universal minimum separation distance: the impact depends on reflector geometry, antenna pattern and required performance.
• Mounting stability: Use a rigid, stable mount. For reference stations, account for monument stability, thermal expansion, wind loading and long-term movement.
• Ground plane and mounting surface: Follow the antenna manufacturer’s specified mounting configuration. A ground plane can improve performance for some antennas, but adding one arbitrarily may alter the pattern and calibration.
• Orientation and reference point: Follow manufacturer and processing requirements for antenna orientation. For high-precision geodetic stations, document the antenna reference point, monument eccentricities, antenna model, radome and orientation consistently in station metadata and observation files.
• Cable, connectors and protection: Use cable with suitable loss at the operating frequencies, weatherproof connectors and appropriate grounding, surge and lightning protection. Keep cable routing and nearby electronics from introducing interference.
• Verification after installation: Inspect satellite C/N0 by elevation and azimuth, tracking continuity, code-minus-carrier or other receiver multipath indicators, measurement residuals and position repeatability. Compare results across time and satellite geometry. No single diagnostic metric proves that multipath is absent.

7. A practical selection workflow

A defensible antenna choice follows the application rather than a generic ranking of antenna types:

• Define the required position or timing performance and the operating environment.
• List the GNSS constellations, frequencies and signals the receiver will use.
• Assess nearby reflectors, sky visibility, platform effects, cable loss and interference risks.
• Shortlist antennas with appropriate radiation patterns, polarisation performance and environmental ratings.
• For precision applications, verify calibration availability and compatibility with the intended processing chain.
• Test the antenna in the actual mounting configuration and assess measurement quality and repeatability.

8. Conclusion

Reliable GNSS performance depends on the receiver, antenna and installation working together. Wider-band signals can improve intrinsic code-multipath resistance, while antenna patterns and polarisation characteristics can reduce some reflected energy. Precision antennas and suitable calibration are particularly important for RTK, surveying and geodetic applications, but calibration does not eliminate site-specific multipath.

The strongest engineering approach is to select an antenna for the actual signals and environment, install it with a clear sky view and stable mounting, apply the correct calibration model, and verify performance using receiver diagnostics and repeatable measurements. Choke-ring antennas are a strong option for many geodetic reference stations, whereas compact patches or other antennas may be more appropriate for size-, weight- or cost-constrained products.

Technical references and further reading

• International GNSS Service (IGS), Antenna Committee and antenna calibration resources. https://www.igs.org/wg/antenna/
• IGS CORS Guidelines, GNSS antenna. https://igs.git-pages.gfz-potsdam.de/igs-cors-guidelines/en/equipment/antenna/
• IGS CORS Guidelines, signal quality and multipath. https://igs.git-pages.gfz-potsdam.de/igs-cors-guidelines/en/establishment/signals/
• NOAA National Geodetic Survey, GNSS antenna calibration. https://nweb.ngs.noaa.gov/research/gnss-satellite/gnss-antenna-calibration.shtml

Lau and Cross (2007), Development and testing of a new ray-tracing approach to GNSS carrier-phase multipath modelling, Journal of Geodesy.

https://doi.org/10.1007/s00190-007-0139-z

For more support contact us at: info@linfinitygnss.com