R&SSMW200A vector signal generator: Testing WLAN 802.11ad signals up to 65 GHz

Generating wideband signals up to 65 GHz for testing WLAN 802.11ad receivers is a major challenge. The R&S®SMW200A vector signal generator in combination with the new R&S®SZU100A I/Q upconverter mastered this task and enables objective evaluation of the performance of these receivers.

Modern wireless communication scenarios require even higher information data rates up to the Gbit class. Large amounts of data are managed in cloud storage and must be quickly available at any time. Wireless devices represent the future: mobile phones that quickly synchronize with multimedia libraries or laptops that exchange large video data flows in 4K quality use wireless docking stations with external hard drives, servers, TVs or projectors.

Developing a signal generator for such applications requires going beyond conventional modulation bandwidths and frequency ranges. The required specifications include the generation of wideband signals up to a 60 GHz frequency range and 2 GHz RF bandwidth with high output power and excellent signal quality. This article describes how the new R&S®SZU100A I/Q upconverter can now extend the capabilities of the R&S®SMW200A vector signal generator all the way to 65 GHz, especially for WLAN 802.11ad applications.

65 GHz – Test and Measurement Challenges

Until now, reproducible testing of mmWave receivers has been very complex. For some time, it has been easy to generate wideband test signals in the frequency range up to 40 GHz directly using vector signal generators such as the R&SSMW200A. However, signals in the 60 GHz range (e.g. for testing WLAN 802.11ad receivers) require additional RF mixers to achieve the target frequencies from 57.32 GHz to 65.80 GHz. But these conventional mixer arrangements have some disadvantages. For example, for practical reasons, upconversion to the 60 GHz band is usually achieved in several stages. This often results in local oscillator (LO) mixing products in the operating band. In addition, filters must be used to suppress the additional sidebands generated during mixing.

Conventional mixing concepts are also subject to fluctuating RF characteristics. Depending on the frequency and level, the device usually has a different frequency response. To cope with this, the actual frequency response must be recorded using external measuring equipment and a tedious correction must be performed. However, this compensation is only valid for one level and one frequency and in practice needs to be recalibrated before each measurement, which results in quite long measurement times.

To complicate matters further, mmWave propagation is significantly more attenuated than frequencies below 6 GHz. For built test setups, 7dB to 10dB of attenuation per meter is expected. However, the tight integration of the antenna array and the RF front end usually does not allow for a wired connection to the WLAN 802.11ad receiver. Therefore, testing WLAN 802.11ad signals can only be done over the air interface.

In a typical over-the-air test setup, the generator signal is delivered to a transmit horn antenna (e.g., with 23dBi gain), and the receiver under test is placed, for example, 1 meter away. The attenuation of a 60GHz signal in free space is 68dB/m, which means that relatively high output power from the signal generator is required to complete these tests. To meet the -53dBm sensitivity limit of a WLAN 802.11ad receiver for an MCS 12 signal, the signal generator must generate a signal with at least -8dBm transmit power. Losses in the test setup caused by switches, adapters, or feeder cables increase the required power to 0dBm or more. If you want to test WLAN receivers to their limits, you must also be able to generate very low levels with the same modulation quality, especially with a good signal-to-noise ratio, which is a major challenge for conventional test setups.

Vector Signal Generator for 65 GHz

A unique solution for making these measurements is the new R&S®SZU100A I/Q upconverter, which extends the R&S®SMW200A vector signal generator to the frequency range of 57.32 GHz to 65.80 GHz (see Figures 1 and 2). The wideband baseband option (R&S®SMW-B9) of this generator generates the internal WLAN 802.11ad signals with the required symbol rate of 1.76 GHz. All WLAN-specific parameters such as modulation, coding, packet size and MAC header can be configured as required. This approximately 2 GHz baseband signal is fed to the R&S®SZU100A via the analog I/Q inputs, where it is upconverted to the 60 GHz band using a local oscillator signal generated by the high-performance RF synthesis module of the R&S®SMW200A. This I/Q upconverter is controlled via USB and seamlessly integrated into the R&S®SMW200A operating concept. Frequencies and levels are usually adjusted via the signal generator’s graphical user interface. Users can operate the unit just as easily and conveniently as any stand-alone vector signal generator from Rohde & Schwarz.

Figure 1: The R&S®SZU100A I/Q upconverter extends the R&S®SMW200A to the 57.32 GHz to 65.80 GHz frequency range for WLAN 802.11ad receiver measurements

Fig. 2: The R&S®SZU100A I/Q upconverter fits seamlessly into the operating concept of the R&S®SMW200A

High output power

The R&S®SZU100A is designed for use as a remote millimeter-wave head. It can be placed close to the device under test and can be flexibly positioned using its adjustable feet and various mounting points. A horn antenna can be mounted directly to its waveguide output (WR15) without the need for an additional adapter. This I/Q upconverter has its own output amplifier, attenuator and integrated level detector mounted directly on the waveguide output. This enables precise level setting from -80 dBm to +5 dBm with excellent linearity over the entire dynamic range and a virtually constant signal-to-noise ratio. Typically, output levels of even greater than +10 dBm can be achieved. This advanced synthesis concept ensures that spurious and images are suppressed and external filters with their undesirable insertion losses are no longer necessary. All these measures significantly reduce the losses in the test setup and ensure that the device under test can be tested with high output powers.

Real-time correction of frequency response

Rohde & Schwarz fully characterizes the I/Q upconverter during production and saves its frequency response correction data in its EEPROM. The R&S®SMW200A uses these values ​​to correct the frequency response in real time during operation (see Figure 3). This ensures a flat frequency response that is independent of level, frequency and signal type, eliminating the need to correct the setup with expensive external calibration hardware before each measurement. This not only reduces hardware costs, but also saves additional calibration cycles during operation and significantly reduces the overall measurement time. Because the frequency response correction is independent of the transmitted signal, it is no longer necessary to save various predistorted waveforms for each individual test setup. The same signal can be used at different measurement sites. This reduces the effort required to manage waveform libraries and provides transparent results. Thanks to the real-time correction of the frequency response, the R&S®SMW200A and R&S®SZU100A combination can achieve the specified excellent error vector magnitude (EVM) value of -31 dB; typically –32 dB or even better can be achieved (see Figure 4).

Figure 3: Frequency response measured with the R&S®SMW200A in combination with the R&S®SZU100A, carrier frequency 64.80 GHz, output level +0 dBm

Figure 4: Measured error vector magnitude (EVM) of a WLAN 802.11ad signal (MCS 12) generated by the R&S®SZU100A and R&S®SMW200A at 60.48 GHz

Summarize

The R&S®SZU100A I/Q upconverter and the R&S®SMW200A vector signal generator are a powerful duo for WLAN 802.11ad measurements in the 60 GHz band. The device generates standard-compliant physical layer signals with excellent signal quality and impressive dynamic range for testing components, modules and wireless devices. Thanks to the internal real-time frequency response correction, the signal is always correct. There is no need for special calibration with additional equipment before each measurement. The device works only with the signal generator: measurements can be started immediately after the necessary configuration. This ensures that the workflow can be set up quickly and easily, allowing users to successfully perform their measurement tasks in the shortest possible time.