Main Content

Modeling and Testing an NR RF Transmitter

The example shows how to characterize the impact of RF impairments, such as in-phase and quadrature (IQ) imbalance, phase noise, and power amplifier (PA) nonlinearities in the performance of a new radio (NR) radio frequency (RF) transmitter. The NR RF transmitter is modeled in Simulink® using 5G Toolbox™ and RF Blockset™.

Introduction

This example shows how to characterize the impact of RF impairments such as IQ imbalance, phase noise, and PA nonlinearities in the performance of an NR RF transmitter. To evaluate the performance, the example considers these measurements:

  • Error vector magnitude (EVM): vector difference at a given time between the ideal (transmitted) signal and the measured (received) signal.

  • Adjacent channel leakage ratio (ACLR): measure of the amount of power leaking into adjacent channels and is defined as the ratio of the filtered mean power centered on the assigned channel frequency to the filtered mean power centered on an adjacent channel frequency.

  • Occupied bandwidth: bandwidth that contains 99% of the total integrated power of the signal, centered on the assigned channel frequency.

  • Channel power: filtered mean power centered on the assigned channel frequency.

  • Complementary cumulative distribution function (CCDF): probability of a signal's instantaneous power to be a level specified above its average power.

The model works on a subframe by subframe basis. For each subframe, the workflow consists of these steps:

  1. Generate the baseband waveform using 5G Toolbox functions.

  2. Upconvert the generated waveform to the passband frequency and apply RF filtering and amplification using RF Blockset.

  3. Downconvert the transmitted waveform to baseband frequency.

  4. Calculate the ACLR/ACPR, occupied bandwidth, channel power, and CCDF using the Spectrum Analyzer block.

  5. Demodulate the waveform at the receiver to measure EVM.

The example uses a Simulink model to perform these operations. Baseband signal processing (steps 1 and 5) uses MATLAB® Function blocks, whereas the RF transmitter modeling (steps 2-4) uses RF Blockset. This model supports Normal and Accelerator simulation modes.

Simulink Model Structure

The model contains three main components:

  • NR Baseband Generation: generates the baseband waveform.

  • RF Transmission: upconverts the waveform from baseband to passband, applies RF filtering, and amplification and performs downconversion.

  • NR Baseband Reception and Measurements: performs the RF measurements and demodulates the baseband waveform to calculate the EVM.

modelName = 'NRModelingAndTestingRFTransmitterModel';
open_system(modelName);

NR Baseband Generation

The NR-TM Transmission block transmits standard-compliant 5G NR test model (NR-TM) waveforms for frequency range 1 (FR1) and frequency range 2 (FR2) [ 1 ].

For the NR-TM waveform generation, you can specify the NR-TM name, the channel bandwidth, the subcarrier spacing (SCS), the duplexing mode, the cell identity, and the TS 38.141 version using the NR-TM Transmission block mask. Additionally, this block provides the option to enable or disable the ACLR test. When the ACLR measurement is enabled, the waveform is oversampled to visualize the spectral regrowth.

For more information on how to generate NR-TMs, see 5G NR-TM and FRC Waveform Generation.

As the example works on a subframe by subframe basis, the NR-TM Transmission block sends one subframe at a time. Transmitting ten subframes, corresponding to one frame in the case of FDD duplexing mode, lasts 10 ms. If the simulation time is longer than 10 ms, the NR-TM Transmission block transmits the same frame cyclically. The Subframe Counter block stores the number of the currently transmitted subframe. If the simulation time is longer than a frame period, the Subframe Counter block resets to 0.

RF Transmission

The RF Transmitter block is based on a superheterodyne transmitter architecture. This architecture upconverts the waveform to the passband frequency and applies RF filtering and amplification before transmitting the signal. Typical components of the superheterodyne transmitter are:

  • IQ modulator consisting of mixers, phase shifter, and local oscillator

  • Bandpass filter

  • Power amplifier

In addition to these components, this RF Transmitter block also includes a variable gain amplifier (VGA) to control the input back-off (IBO) level of the high power amplifier (HPA).

set_param(modelName,'Open','off');
set_param([modelName '/RF Transmitter'],'Open','on');

The Inport block converts the complex baseband waveform into an RF signal and the Outport block converts the RF signal back into complex baseband. Because the RF Transmitter accepts a maximum of 1024 samples per subframe, the Input Buffer, before the RF Transmitter block, reduces the number of samples sent to the RF Transmitter. In the current configuration, the Input Buffer sends 1024 samples at a time.

Before sending the samples onto the Decode Subframe block, the Output Buffer (after the RF Transmitter) buffers all samples within a subframe.

The Delay block accounts for buffer-induced delays. As the duration of the delay is equivalent to the transmission of a subframe, the Decode Subframe block does not demodulate the first information received during a subframe period.

You can configure the RF Transmitter components using the RF Transmitter block mask.

The RF Transmitter block exhibits typical impairments, including:

  • I/Q imbalance as a result of gain or phase mismatches between the parallel sections of the transmitter chain dealing with the IQ signal paths.

  • Phase noise as a secondary effect directly related to the thermal noise within the active devices of the oscillator.

  • HPA nonlinearities due to DC power limitation when the amplifier works in saturation region.

This example highlights the effect of nonlinear behavior of the HPA.

NR Baseband Reception and Measurements

The Decode Subframe block performs OFDM demodulation of the received subframe, channel estimation, and equalization to recover and plot the PDSCH symbols in the Constellation Diagram. This block also averages the EVM over time and frequency and plots these values:

  • EVM per OFDM symbol: EVM averaged over each OFDM symbol.

  • EVM per slot: EVM averaged over the allocated PDSCH symbols within a slot.

  • EVM per subcarrier: EVM averaged over the allocated PDSCH symbols within a subcarrier.

  • Overall EVM: EVM averaged over the allocated PDSCH symbols transmitted.

According to TS 38.141-1 [ 1 ], not all PDSCH symbols are considered for the EVM evaluation. Using the RNTI, the helper function hListTargetPDSCHs selects the target PDSCH symbols to analyze.

The Spectrum Analyzer block provides frequency-domain measurements such as ACLR (referred to as ACPR), occupied bandwidth, channel power, and CCDF. To visualize the spectral regrowth, the ACLR test oversamples the waveform. The oversampling factor depends on the waveform configuration and should be set such that the generated signal is capable of representing first and second adjacent channels. The ACLR evaluation follows the specifications in TS 38.141-1 [ 1 ].

Model Performance

To characterize the impact of HPA nonlinearities in the EVM and ACLR evaluations, you can measure the amplitude-to-amplitude modulation (AM/AM) of the HPA. The AM/AM refers to the output power levels in terms of the input power levels. The helper function hPlotHPACurve displays the AM/AM characteristic of the HPA selected for this model.

hPlotHPACurve();
figHPA = gcf;

P1dB is the power at 1 dB compression point and is usually used as a reference when selecting the IBO level of the HPA. You can see the HPA impact on the RF transmitter by analyzing the EVM and ACLR results for different operating points of the HPA. For example, compare the case when IBO = 15 dB, corresponding to HPA operating in linear region, with the case when IBO = 2 dB, corresponding to HPA operating in full saturation. The gain of the VGA controls the IBO level. To keep a VGA linear behavior, select gain values lower than 20 dB.

  • Linear HPA (IBO = 15 dB). To operate at an IBO level of 15 dB, set the Available power gain parameter of the VGA block to 0 dB. To simulate a whole frame, run the simulation long enough to capture 10 subframes (10 ms). During simulation, the model displays the EVM and ACLR measurements and the constellation diagram.

set_param([modelName '/RF Transmitter'],'vgaGain','0');
sim(modelName);
--- Starting simulation ---
 Transmitting subframe 0 ...
 Transmitting subframe 1 ...
 Transmitting subframe 2 ...
 Transmitting subframe 3 ...
 Transmitting subframe 4 ...
 Transmitting subframe 5 ...
 Transmitting subframe 6 ...
 Transmitting subframe 7 ...
 Transmitting subframe 8 ...
 Transmitting subframe 9 ...

--- End of simulation ---

According to TS 38.104 [ 2 ], the minimum required ACLR for conducted measurements is 45 dB and the maximum required EVM when the constellation is 64-QAM is 8%. As the ACLR values are higher than 45 dB, and the overall EVM, around 1.3%, is lower than 8%, both measurements fall within the requirements.

  • Nonlinear HPA (IBO = 2 dB). To operate at an IBO level of 2 dB, set the Available power gain parameter of the VGA block to 12 dB.

set_param([modelName '/RF Transmitter'],'vgaGain','12');
sim(modelName);
slmsgviewer.DeleteInstance();

% Restore to default parameters
set_param([modelName '/RF Transmitter'],'vgaGain','0');
--- Starting simulation ---
 Transmitting subframe 0 ...
 Transmitting subframe 1 ...
 Transmitting subframe 2 ...
 Transmitting subframe 3 ...
 Transmitting subframe 4 ...
 Transmitting subframe 5 ...
 Transmitting subframe 6 ...
 Transmitting subframe 7 ...
 Transmitting subframe 8 ...
 Transmitting subframe 9 ...

--- End of simulation ---

Compared to the previous case, the constellation diagram is distorted and the spectral regrowth is higher. In terms of the measurements, the first adjacent channel ACLR does not fall within the requirements of TS 38.104 [ 2 ] and the overall EVM, around 7%, is higher.

The RF Transmitter is configured to work with the default values of the NR-TM Transmission block and with a carrier centered at 2140 MHz (FR1). If you change the carrier frequency or the values in the Waveform Parameters block, you may need to update the parameters of the RF Transmitter components as these parameters have been selected to work for the default configuration of the example. For more information, see the Summary and Further Exploration section of this example.

Summary and Further Exploration

This example demonstrates how to model and test an NR RF transmitter in Simulink. The RF transmitter consists of an IQ modulator, a bandpass filter and amplifiers. To evaluate the performance, the Simulink model considers ACLR and EVM measurements. The example highlights the effect of HPA nonlinearities on the performance of the RF Transmitter. You can explore the impact of altering other impairments as well. For example:

  • Increase I/Q imbalance by using the I/Q gain mismatch (dB) and I/Q phase mismatch (Deg) parameters on the IQ Modulator tab of the RF Transmitter block.

  • Increase the phase noise by using Phase noise offset (Hz) and Phase noise level (dBc/Hz) parameters on the IQ Modulator tab of the RF Transmitter block.

Additionally, you can check the occupied bandwidth, the channel power, and the CCDF measurements by using the Spectrum Analyzer block.

If you change the carrier frequency or the values in the Waveform Parameters block, you may need to update the parameters of the RF Transmitter components as these parameters have been selected to work for the default configuration of the example. For instance, a change in the carrier frequency requires revising the bandwidth of the filter. If you select a bandwidth wider than 20MHz, you may need to update the Impulse response duration and Phase noise frequency offset (Hz) parameters of the IQ Modulator block. The phase noise offset determines the lower limit of the impulse response duration. If the phase noise frequency offset resolution is too high for a given impulse response duration, a warning message appears, specifying the minimum duration suitable for the required resolution. For more information, see IQ Modulator (RF Blockset).

This example could be the basis for testing NR-TM waveforms for different RF configurations. You can try replacing the RF Transmitter block by another RF subsystem of your choice and configuring the model accordingly.

References

  1. 3GPP TS 38.141-1. "NR; Base Station (BS) conformance testing Part 1: Conducted conformance testing." 3rd Generation Partnership Project; Technical Specification Group Radio Access Network.

  2. 3GPP TS 38.104. "NR; Base Station (BS) radio transmission and reception." 3rd Generation Partnership Project; Technical Specification Group Radio Access Network.

Related Topics