As recommended by the 5GAA VATM TR, the maximum elevation of the gantry, should be at least 120 degrees. To minimize test system size, anechoic chambers are designed so that the arm comes close to the floor absorbers at this maximum elevation. Lifting the car enables measurements down to this elevation angle with the exact height determined from the measurement range length. As an example, if the range length from the center coordinates at the base of the car to the tip of the measurement antenna is 6 m, then going down to 120 degree elevation requires a height of at least 6 m times cos(60 degrees), or 3 m.
Achieving the proper distance from the test probe to the absorbers may require additional height. Continuing the previous example and assuming that the test system measures down to 400 MHz, 1.5 m must be added to the previously calculated 3 m height. This means that the DUT has to be raised above the floor by about 4.5 m to provide absorber tips with enough distance at the maximum elevation range. The gantry also lifts similarly, as the center of rotation of the arm should remain in the same plane as the center of coordinates, the vehicle center projected to the turntable surface.
Measurements down to 120 degrees in elevation require specific care in DUT fixturing. A vehicle can weigh several tons, so a large metal structure is required at the azimuth pole to position, lift and rotate the car. Metal creates scattering and perturbates measurements, so absorber panels typically hide the related structures. However, these panels create shadowing effects and absorb waves that should be measured. To reduce these effects, fixtures minimize the footprint extending beyond the vehicle. The 3D model in Figure 4 shows this approach with only small pieces of the fixture visible next to the wheels.
WHY MEASURE UNDER THE FLOOR?
Figure 5 Angle of arrival for cellular technologies.
Cars are used on the road, so it may appear absurd to measure the radiation patterns down to “30 degrees below the floor.” This choice is justified by practical scenarios that are detailed in the 5GAA VATM. As one example, Figure 5 shows communication with a mobile network. Evaluating the cellular propagation channels, the most significant angles of arrival linking a UE to a base station are from 60 to 90 degrees in elevation. However, because of ground reflections, signals from 90 to 120 degrees of elevation are also relevant in an anechoic environment, particularly to transceivers located on the sides of the car. The antenna radiation performance in those directions impacts connectivity and must be qualified.
Satellite communications present another interesting use case. The region of interest is in a cone looking at elevation angles towards the zenith (0 to 60 degrees), as waves are coming from the sky. However, good GNSS service performance requires a good signal-to-noise ratio, so measuring the spatial selectivity of the antennas and filtering out angles of arrival between 60 and 110 degrees is critical to minimize the power captured from poor-quality links coming at grazing angles. This is also addressed in the 5GAA VATM.
Addressing these considerations, the 5GAA VATM defines associated OTA metrics with the partial radiated power (PRP) or partial isotropic sensitivity (PIS). These quantities reflect the overall radiated power or sensitivity performance of the DUT in the relevant angular regions. As an example, PRP is calculated in Equation 1 and Equation 2:
Where:
i=0 to p, is the index of an elevation cut in the EIRP patterns
cuti results from discretized integration of the total EIRP including azimuth and elevation polarizations (θ, φ) in the θi-plane.
p and q are the total number for elevation and azimuth angle intervals, respectively. These are calculated based on the defined angle range of interest and the step size. The PRP is then the cumulated cuti powers in the cuts where the elevation angles sweep through the angular region of interest. In the case of communicating to the mobile network, PRP and PIS, integrated through the 60 to 120 degree elevation region, would be the quantities of interest.
DEALING WITH LIMITED RANGE LENGTH
One of the concerns to address in full-vehicle OTA measurements is the size of the DUT. OTA quantities should be measured in the far-field. Far-field is, ideally, captured at infinity, but a practical and common approach is to select a measurement range length, R, greater than the Rayleigh or Fraunhofer distance (FHD). This reasonably approximates the far-field condition and is shown in Equation 3:
Where:
D is the antenna radiation aperture
λ is the free-space wavelength.
Figure 6 Magnitude of electric currents at 1.8 GHz.
However, except in canonical cases, the exact antenna aperture is not known a priori. Since antennas couple with nearby structures, the aperture will generally be larger than the antenna element itself. At a minimum, D will not be larger than the diameter of the minimum sphere encompassing the whole DUT. This is a workaround that can be used with compact devices like smartphones. The rationale behind the worst-case range length definition for CTIA OTA testing is based on considering D as the maximum DUT dimension. However, a large, high-end car with a 6 m length, supporting 5G NR FR1 communication up to 7.125 GHz returns a calculated D value of 6 m and an R of more than 1.7 km. Clearly, full-vehicle OTA testing requires more advanced considerations.
A car is a large device, but the integrated antennas are very localized. The car does not behave as one big antenna. Even though its structure couples to the radiated fields from the antenna elements, the currents spread over a limited region. As an example, the EMPIRE XPU simulation presented in Figure 6 shows the magnitude of electric currents over the conducting car roof when the shark fin antenna from Figure 2a is excited with a 1.8 GHz signal. The amplitude decays quickly, down by 20 dB within 20 cm from the feed point, which is a little more than a wavelength. Standalone shark fin antenna evaluations typically use round metallic plates to simulate these close energy-coupling effects, but using size-limited ground planes, typically a 1 m diameter disk, for measurements down to 600 MHz means only one wavelength radius at this frequency.
Figure 7 Far-field error boundary isolines.
Also, as demonstrated in the past years, FHD may be overkill for certain OTA tests. If a certain error on the measured radiated metrics can be tolerated, then a shorter effective far-field distance can be used. Detailed considerations about this distance can be found in the related ANSI C63 white paper.5 Figure 7 continues the 6 m range length gantry system example and shows far-field error boundaries as a function of D and frequency. This plot shows that a 15 percent deviation, less than 0.7 dB, allows far-field OTA test antenna apertures as large as 75 cm at 8 GHz and 1.8 m at 500 MHz. Combining this finding with the considerations of Figure 6 gives confidence that 5GAA VATM-type ranges allow true far-field OTA measurements of vehicle-integrated transceivers.
When performing measurements at shorter range lengths, the offset between the center of radiation, typically at the antenna element and the center of the coordinate system must be considered. Ignoring this offset can result in large errors in the measured radiated quantities. To mitigate these effects, the car can be physically translated in the test environment by adding one or more linear axes at the turntable or processing the test signals for receive and transmit measurements.6
Another alternative from the 5GAA VATM to address the limited range length involves combining passive and active measurements.3 In this scenario, the fields radiated by the antennas in the DUT are first characterized in magnitude and phase with a vector network analyzer. Near-field to far-field processing is then applied to deduce the actual far-field directivity. Finally, an OTA test is run with the DUT in the same position, utilizing the same DUT antenna, but this time cabled to the transceiver and transmitting or receiving digitally-modulated signals. The setup change does not affect the propagation of electromagnetic fields from the measured scan, assumed to be in the near-field to the far-field region. The ratio of power in a given direction between the scanned sphere and a sphere at a larger distance obtained in the passive antenna test can be used to derive the actual far-field OTA values based on the near-field OTA scan.
CONCLUSION
Qualifying the connectivity performance of embedded car modules requires system-level OTA characterization that includes the vehicle. Without international standards covering these measurements, the 5GAA VATM TR has become the de facto standard, particularly to define test environments. Systems and software supporting CTIA’s “Certification Test Plan for Wireless Over-the-Air Performance” are appropriate to execute the related measurements. This paper has provided some implementation details on the industry-preferred methodology using a large anechoic chamber, including a distributed-axis spherical scanner with an elevation range of up to 120 degrees. It has shown that full-vehicle far-field OTA tests can be made in compact 5GAA VATM-compatible environments. Finally, techniques were introduced to enable accurate measurements by mitigating the effects of limited measurement range length.
ACKNOWLEDGMENT
The authors would like to thank Winfried Simon at IMST GmbH for the vehicle antenna simulation and figure support using the EMPIRE XPU software.
References
- A. Lauer, W. Simon and A. Wien, “XPU Technology for Fast and Efficient FDTD Simulations using Modern CPUs Cache Memory Bandwidth,” European Conference on Antennas and Propagation, Vol. 13, No. 15, pp. 2584–2589, March 2015.
- CTIA Certification Test Plan for Wireless Device Over-The-Air Performance V4.0.X, Web: https://ctiacertification.org/test-plans/.
- 5GAA Technical Report, “Vehicular Antenna Test Methodology,” Aug. 2021, Web: https://5gaa.org/vehicular-antenna-test-methodology/.
- GB/T XXXX-XXXX draft standard, Road vehicles – RF Performance Requirements and Test Methods for Vehicle Antenna Systems, Jan. 24.
- ANSI C63, Discussion on Measurement Test Distance for Determining EIRP or TRP of Active Antenna Systems, accepted for publication.
- G. F. Hamberger, et al., “Correction of Over-the-air Transmit and Receive Wireless Device Performance Errors Due to Displaced Antenna Positions in the Measurement Coordinate System,” IEEE Trans. Antennas Propag., Vol. 68, No. 11, June 2020, pp. 7549–7554.
