5G Focuses on OTA

In 5G, highly integrated massive MIMO antenna systems that support beamforming are superseding classic sector antennas. These systems must be characterized over the air – a challenge for T&M manufacturers and system developers alike. Achieving the needed high channel capacity in a 5G network requires rolling out massive MIMO base stations along with networks and mobile stations implementing both microwave and millimeter wave technologies. In 5G, the microwave range is designated as frequency range 1 (FR1, 410 MHz to 7.125 GHz) and the millimeter wave range as frequency range 2 (FR2, 24.25 GHz to 52.6 GHz).

In FR1, the main innovation effort is on the base station (BS). Active arrays (massive MIMO arrays) are deployed that can have hundreds of antenna elements. Massive MIMO has two objectives. First, it allows the generation of multiple independent data streams for simultaneous coverage of multiple mobile stations (multi-user or MU-MIMO). Second, beamforming can be used to direct these streams at specific remote stations. This focuses the energy in order to increase the coverage area, reduce interference and boost the data rate. There is also a bonus side effect of lower energy consumption, which reduces overall network costs.

In FR2, the transmission systems use large available bandwidths at frequencies around 28 GHz and 39 GHz. These high frequencies lead to high path loss based on the formula F = (4πrf/c)2 and large electromagnetic field absorption in nearby objects. Array antennas with their improved gain help reduce this loss.

Performance Metrics Quantified Over-The-Air (OTA)

In FR2, the need for low path losses and small dimensions is leading to highly integrated solutions combining antennas, modems, amplifiers and phase shifters. Consequently, there are no longer any RF contacts for connecting test instruments via cables. This is why over-the-air (OTA) test solutions are needed in order to characterize transmit and receive antennas. Such solutions are also needed to verify whether the transmission lobe is pointed in the right direction.

Figure 1: The R&S ATS1000 is an example of a spherical measuring system that supports both direct far-field measurements and software-based near-field to far-field transformation. The DUT can be precisely positioned with the aid of a laser. The system can be equipped with a hemispherical climatic chamber that fully encloses the DUT and allows measurements under climatic stress across a wide temperature range.

The metrics to be determined include antenna parameters for the radiated and received power such as the effective isotropic radiated power (EIRP), total radiated power (TRP), effective isotropic sensitivity (EIS) and total isotropic sensitivity (TIS). There are also transmitter-specific metrics such as the error vector magnitude (EVM), adjacent channel leakage ratio (ACLR) and spectrum emission mask (SEM).

Antenna characteristics are generally measured in the homogeneous far field (FF). However, far-field conditions at FR2 frequencies for base station antenna sizes do not begin until a distance of many meters from the antenna. Using direct FF probing and applying the Fraunhofer distance (FHD) criterion for the far field (r = 2D²/λ, D is the antenna aperture), a 75 cm massive MIMO device under test (DUT) radiating at 2.4 GHz would require a test chamber with a range of at least 9 m.

Even a 15 cm smartphone transmitting at 43.5 GHz would require a testing distance of 6.5 m. This distance is required to create a region encompassing the DUT, the quiet zone (QZ), with the necessary phase constancy. The impinging field must be as uniform as possible and approach a plane wave with phase deviation below 22.5 °.

Theoretical research shows that actual FF behavior in the peak directivity region can start much closer in than the FHD. Research results proved, for example, that the FF EIRP or EIS of a 15 cm DUT radiating at 24 GHz can be accurately determined at a distance as short as 1.14 m. This distance reduction of about 70 % is achieved at the cost of increased longitudinal taper error. Also, side lobe levels cannot be evaluated accurately at shorter distances.

While direct FF measurements at shorter distances are not practical for all applications, they are beneficial in cases where the necessary conditions are met. This is because large anechoic chambers have high costs of ownership and limited dynamic range due to the path losses. The simplest scenario is the “white box” case where the antenna location within the device and its aperture size are known and the aperture fits entirely within the QZ.

If this is not the case or if the DUT has multiple antennas situated on opposite edges of the enclosure, then we have the “black box” scenario where the radiating currents can flow anywhere within the DUT. Due to the potentially larger aperture, a significantly larger QZ must be implemented, making direct FF measurements difficult. For such situations, we can employ near-field to far-field transformations (NFFF) as an alternative.

Figure 2: The reflector’s shape and manufacturing precision are important quality parameters in a CATR system. The R&S ATS800B CATR benchtop system is a cost-effective solution for development and research laboratories that need high-quality measurement results.

Software NFFF Transformations

A first efficient approach for reducing the FF distance (and thus the necessary size of the shielded chamber) involves the use of software transformation methods. The mathematical implementations may vary, but the concept is generally the same: at least two polarization components of the electromagnetic field (E, H or a mixture of the two) are measured in magnitude and phase over a spherical surface encompassing the DUT.

The measured data is processed using mathematical functions to propagate the fields towards larger distances and extract far-field radiation components. From the Huygens principle, the knowledge of two phasors (complex amplitudes) is enough to reconstruct exactly all six field components outside the surface. Alternative transformation methods use spherical wave expansion (SWE), plane wave expansion (PWE) or integral equation resolution, along with techniques to improve computational efficiency or accuracy by taking parameters such as the spatial sampling rate, scanning area or truncation into account.

Fig. 1 shows a system capable of measurements using spherical scanning around the DUT. The DUT is positioned on a turntable rotating in azimuth. A dual-polarized Vivaldi antenna is mounted at the tip of a boom rotating in elevation. The DUT is connected to one port of a vector network analyzer (VNA). The antenna ports for the two polarization planes are connected to two other VNA ports, enabling measurement of complex S-parameters such as the transmission and reflection coefficient.

Near-field measurement methods often rely on assumptions applying to this above-described case of “passive or RF-fed antenna testing”:

• The antenna feed port is accessible with a signal fed to the antenna that is used as phase reference
• The RF signal is a continuous wave (CW) signal
• Reciprocity applies so that transmit (TX) and receive (RX) patterns are identical at the same frequency

There are workarounds available in TX cases where such assumptions do not apply. Hardware and processing implementations to retrieve the propagation phase vary, for example using interferometric techniques or multi-port phase-coherent receivers, with the addition of a dedicated phase reference antenna that has to be installed in the vicinity of the radiating DUT. Alternative approaches include phaseless methods where the phase information is retrieved from magnitude-only measurements.

The RX case is more complex because usually the entire RX chain must be measured since the sensitivity is verified based on attainment of the minimum required data throughput. The reciprocity assumption does not apply since the RX RF component chain is in general different from the TX RF chain. Furthermore, for a receiving DUT with no antenna test port, the power available at the input to the RF frontend cannot be straightforwardly predicted. There is also no access to a phase reference in this case so that the software NFFF transformation becomes inapplicable. Therefore, effective isotropic radiated power (EIRP) can be evaluated accurately in the near field using software NFFF but not effective isotropic sensitivity (EIS).

Figure 3: The R&S ATS1800C test chamber is used for (pre)conformance testing of reference designs and mobile devices. The positioner can accommodate DUTs with sizes up to about DIN A4 and weighing up to several kilograms.

Software NFFF methods reach their limits when attempting to determine transceiver metrics such as EVM, ACLR and SEM. Such information must be extracted directly from the modulated signal. However, software NFFF solutions only process the complex amplitude values used to derive a three-dimensional representation of the antenna characteristics. Using advanced T&M methods, however, this is not a major limitation since compact measuring systems can create an “indirect far field”, allowing measurements to be performed that are comparable to the true far field.

Hardware-Based Near-Field to Far-Field Transformation Provides Clarity

Different testing methods enable direct OTA assessment close to the antenna without applying a software transformation. In these hardware approaches, the idea is to physically create far-field conditions in a specified quiet zone (QZ) region within a short range. This “indirect far field” can be created in a compact antenna test range (CATR) or by using plane wave synthesis.

Compact Antenna Test Range

A CATR uses a parabolic reflector to transform a spherical wave-front from the DUT into a planar wave-front (Figs. 2 and 4). The quality of the measurement results that can be obtained with a CATR depends on the reflector quality. The edge shape and surface roughness influence the frequency range in which an acceptable quality quiet zone can be produced. The edge configuration limits the lowest operating frequency while the surface roughness determines the upper frequency. Serrated or rolled edges prevent diffraction that could otherwise significantly contaminate the QZ. The reflector size with serrated/rolled edges is generally at least two times the DUT/QZ size whereas a reflector with sharp edges is three to four times the size of the QZ.

The feed antenna’s radiation pattern has a direct impact on the size of the QZ since the reflector essentially “projects” the pattern onto the QZ. Since the QZ size depends more on the reflector characteristics than on how far the DUT is from the reflector, it is much easier to create a large QZ inside small enclosures, which makes testing simpler. The CATR test setup shown in Figs. 2 and 4 fits inside a chamber measuring only 2.1 m × 0.8 m × 1 m (R&S ATS800R). A direct FF measuring system would require a range of up to 14.5 m.

Figure 4: Compact antenna test range (CATR) with a roll-edged reflector collimating a spherical wavefront into a planar wavefront (fields computed with a model of the actual setup implemented in the CST MWS simulation software at 28 GHz). 

CATRs are of great interest for testing mobile and base stations operating in 5G NR FR2 since they considerably reduce the size of the test environment and produce measurement results directly, i.e. without any further NFFF computations. In addition, CATRs have the same capabilities as an FF system in terms of direct measurements of RF transceiver metrics in both TX and RX mode. Finally, since the path loss of such a system only occurs in the region between the feed and the reflector, the dynamic range of a CATR system is also improved compared to a direct FF approach.

Plane Wave Synthesis Using a Phased Antenna Array

While CATRs can be built for 5G millimeter wave DUTs using relatively compact and lightweight reflectors (20 kg to 40 kg), the reflector weight increases significantly in FR1 – up to hundreds of kilograms for DUTs the size of base stations. The cost, fabrication time and handling of the large, heavy reflectors become prohibitive. An “electronic version” offers a lightweight, cost-effective alternative. It consists of an antenna array with multiple elements that are individually controlled in amplitude and phase to produce a planar wave-front starting at a relatively short distance.

Rohde & Schwarz has developed such a planar wave converter (PWC) consisting of 156 wideband Vivaldi antennas and a network of phase shifters and attenuators. This PWC array is 1.7 m wide and creates a spherical QZ of 1 m diameter at a distance as short as 1.5 m in the frequency range from 2.3 GHz to 3.8 GHz (Fig. 5). A combined-axis positioner is used to position the DUT (e.g. a base station antenna) in the QZ. In the figure, the DUT is a calibration antenna that is used to control the levels of individual RF channels as well as to determine the path loss of the entire test system. The PWC is reciprocal and has only one RF input/output, which can be connected to a signal generator, spectrum analyzer or vector network analyzer.

Summary: OTA Measurements are Simpler Than Ever

The need to test 5G components over the air requires new, more sophisticated measuring equipment than previous mobile communications technologies. The challenge for T&M manufacturers is producing such equipment at competitive prices while ensuring simple operation. The solution is a measuring system that can reliably assess the behavior of a DUT in the far field without having to install large shielded chambers that satisfy the Fraunhofer far-field criterion. Near-field techniques employing software transformations are suitable for evaluation of EIRP and TRP quantities.

Figure 5: The heart of the plane wave converter measuring system from Rohde & Schwarz is the R&S PWC200 phased array (left). In place of the DUT is a calibration antenna array mounted on a great circle cut positioner.

When RX or demodulation is involved with a DUT using multiple non-identical RF transceivers, methods utilizing hardware field transformations such as CATR and PWC can overcome the limitations of software NFFF. They also provide compact and reliable alternatives to direct far-field measurements, putting them in pole position for future 3GPP RF conformance testing of mobile devices and base stations.