Industry Insights

Overcoming 5G NR mmWave Signal Quality Challenges

2019-09-30  |  9 min read 

Many factors impact signal quality including baseband signal processing, modulation, filtering, and up conversion. Common signal impairments become more problematic at higher frequencies or with wider bandwidths. With wider channel bandwidths at millimeter-wave (mmWave) frequencies, these impairments impact baseband and RF designs.

Inherent in orthogonal frequency-division multiplexing (OFDM) systems, orthogonal properties prevent interferences between overlapping carriers. However, impairments such as IQ impairments, phase noise, linear and nonlinear compression, and frequency error cause distortion in the modulated signal. Phase noise is one of the most challenging factors in mmWave OFDM systems. Too much phase noise in designs results in each subcarrier interfering with other subcarriers leading to impaired demodulation performance.

Such issues impact the performance of your designs and are difficult to resolve. Device designs need to overcome the physical challenges in wide bandwidth and mmWave signals. Test solutions require better performance than the device under test (DUT) to properly measure and characterize the signal quality without introducing new issues.

Characterizing Signal Quality

Evaluating a signal’s modulation properties provides one of the most useful indicators of signal quality. Viewing the IQ constellation helps in determining and troubleshooting distortion errors. Another key indicator of a signal’s modulation quality is a numeric error vector magnitude (EVM) measurement that provides an overall indication of waveform distortion.

5G NR specifies a cyclic prefix OFDM (CP-OFDM), which is a multi-carrier modulation scheme. An EVM measurement reflects any variation in a circuit’s phase, amplitude, or noise seen in wideband signals. EVM is the normalized ratio of the difference between two vectors: IQ measured signal and IQ reference (IQ reference is a calculated value) as shown in figure 1. EVM is a measure of  the average amplitude of the error vector from the ideal reference point. It is typically measured in dB or as a percentage.

Figure 1 visualizes the EVM calculation.

EVM calculation

With the expected use of higher-order modulation schemes in 5G (up to 256 QAM initially, and up to 1024 QAM in the future), components and devices require a better EVM result as the modulation density increases. Table 1 shows how 3GPP EVM requirements for user equipment (UE) get tighter as the modulation density increases.

Table 1 shows the 3GPP TS 38.101-1 EVM requirements for different 5G modulation schemes.

Modulation scheme for PDSCH

Requires EVM






16 QAM




64 QAM




256 QAM



Spectrum measurements are also necessary to validate a signal’s RF performance. 5G UE spectrum measurements for transmitting products include measurements such as transmitted power, occupied bandwidth (OBW), adjacent channel power ratio (ACPR), spectrum emissions masks (SEM), and spurious emissions.

A test solution needs to have enough performance to evaluate the constellation diagram and measure the EVM required by 5G components and devices. Flexibility to make spectrum measurements and scale to higher frequencies and bandwidths is required as the 5G standards evolve.

Defining a Measurement Solution

To achieve high quality measurements of high bandwidth devices at mmWave frequencies requires a test solution with EVM performance that is better than the product or system under test. Typical guidelines to follow include:

•            For component test: 10 dB better than the system as a whole

•            For system test: 3 dB better than the source from the radio standard

When measuring a transmitter, receiver, transceiver, or other component in a wireless device, a test solution typically consists of a stimulus and DUT, a DUT and analyzer, or a stimulus, DUT and analyzer, depending on the DUT. Measurements in baseband and sub-6 GHz can typically be conducted using cables. Measurements at centimeter-wave or mmWave frequencies, however, likely require an over-the-air (OTA) measurement due to the high level of integration in the antennas and radio frequency integrated circuits (RFICs) resulting in no connector test points for conducted test.

Figure 2 shows the 5G R&D Test Bed test setup. It has the performance needed to evaluate 5G components and devices for impairments at mmWave frequencies. A vector signal generator produces a digitally modulated 5G NR signal into the DUT. A vector signal analyzer captures the RF signal properties out of the DUT and digitize the modulated signal for analysis. This test solution offers flexible configurations to address the many combinations of frequency, bandwidth, and fidelity required for testing 5G components and devices.

Figure 2 depicts the 5G R&D Test Bed with 5G NR hardware and software.

The test setup itself can introduce other sources of error in a measurement system. When considering a test setup at higher frequencies with wider bandwidths, items such as test fixtures, cables, adaptors, couplers, filters, preamplifiers, splitters, and switching between the DUT and measurement equipment have greater impact than in sub-6 GHz measurement systems. Calibrating the measurement system to the reference plane at the location of the DUT is essential to achieve the highest measurement accuracy. The goal is to see the true characteristics of the DUT without seeing the impacts of the test setup. The measurement system needs to perform better than the DUT design goals. Measurements at the DUT plane provides better measurement accuracy and repeatability. A proper system level calibration eliminates uncertainties due to test fixtures in frequency and phase and is valuable for very wide bandwidth signals. The 5G R&D Test Bed solution includes the Signal Optimizer software that moves the calibration plan from the test equipment to the DUT reference plane as shown in figure 2.

A 5G NR mmWave Measurement

Proper selection of test equipment, connectors, adapters, and system level calibration enables high-performance measurements to evaluate the true performance of a 5G component or device. Figure 3 shows a calibrated measurement of a 5G antenna using Keysight’s 5G R&D Test Bed solution that enables characterization of 5G NR devices from RF to mmWave frequencies with precision and modulation bandwidths up to 2 GHz. With 5G NR compliant software, waveforms are easily created and analyzed with 5G numerology, uplink, and downlink, to test 5G NR and LTE integration and coexistence.

Figure 3 shows the analysis of a 5G NR 256 QAM signal with antenna pattern.

Analysis of 5G NR 256 QAM signal

5G mmWave presents a significant challenge for participants across the 5G mobile ecosystem. For more information on solutions to overcome this challenge, visit the following webpages: