Redefining RF and Microwave Instrumentation Through Open Software and Modular Hardware

Publish Date: Jun 11, 2012 | 13 Ratings | 4.31 out of 5 | Print

Table of Contents

  1. The Spectrum Analyzer as a Case Study
  2. Crucial Elements of a Redefined RF System
  3. Examples of the Innovation of Redefined RF Instrumentation
  4. Instrumentation that Stands the Test of Time

If you are an engineer making measurements or performing automated tests on electronic devices, you either work with or will work with RF and microwave measurements—there’s no denying that reality. Wireless communication has permeated almost every electronic device, from refrigerators and electrical meters to the more commonly acknowledged mobile phones and tablet PCs. But when the time comes to add RF to a test system, it may prove significantly more challenging than adding other measurements of the past.

Putting aside the high-frequency measurement theory (which can be learned with enough effort), the business challenges can quickly sideline the competitiveness of the wireless capability in your latest device. First, RF/microwave instrumentation can easily be the most expensive part of any test and measurement system from a capital equipment standpoint, and the time associated with making these measurements rises quickly if the equipment isn’t optimized. Second, rapidly changing commercial standards such as LTE, 802.11ac, Bluetooth, and so on make a continual learning process mandatory, and more standards have been added than subtracted, so test times keep increasing. Finally, traditional test equipment vendors have largely deployed a paradigm of single-protocol or at least limited-protocol instrumentation that creates discontinuities in even the most well-crafted system architecture.

This creates a need for more openness in software and more modularity in hardware for RF and microwave instrumentation. To meet this need, National Instruments has redefined the traditional approach by combining PXI hardware and NI LabVIEW system design software. The company also is leveraging commercial technologies such as multicore microprocessors, user-programmable FPGAs, PCI Express hardware, and system design software to meet this flexibility and scalability demand for future high-frequency test and measurement applications.

1. The Spectrum Analyzer as a Case Study

To understand how this transition has occurred, consider the classic spectrum analyzer and its fixed, inefficient architecture. The traditional spectrum analyzer uses a swept-tuned approach to spectrum measurements: the local oscillator is swept through the band of interest, and the resulting intermediate frequency signal goes through an analog resolution bandwidth (RBW) filter. This increases test times and reduces flexibility.

If one of your concerns as a test engineer is test times for a spectrum, the time associated with making these measurements is essentially dependent on the sweep speed of the local oscillator (see Figure 1). The process of moving the data to the PC for analysis or data logging has little effect on measurement speed at that point because RF measurements usually take an order of magnitude longer (tens to hundreds of milliseconds) than the milliseconds of latency introduced by controlling an instrument over LAN or GPIB. So, you weren’t focused on the communication bus as a means to speed things up because the measurement architecture made it futile.

Traditionally, if you wanted to scale the capability of the spectrum analyzer for better standards coverage, you focused on the vendor. Much of the instrument’s functionality was effectively set through the instrument vendor’s choice of analog component and fixed signal processing to accommodate the greatest number of use cases when the instrument was developed. This is an older paradigm that has not adapted to today’s wireless needs. If you followed it, you’d find yourself buying additional variants of the analyzer to test new standards.

Figure 1. The traditional spectrum analyzer architecture has fixed analog components and measurement speeds determined by the sweep speed of the local oscillator.

Today, most RF signal analyzers, like NI PXI RF signal analyzers, use a digital approach. Digital RF signal analyzers generally perform spectrum measurements significantly faster than older swept-tune spectrum analyzers. By moving more of the instrument into the digital domain, advances in software architectures as well as processing power derived from Moore’s law can significantly improve test times and flexibility.

Unlike the traditional spectrum analyzer, measurement speed in digital signal analyzers is usually limited by the performance of internal components such as digital preprocessing blocks (digital downconverter (DDC) on the FPGA), the data transfer bus (PCI Express), the processing engine (multicore central processing units), and signal processing algorithms (optimized for multicore CPUs by LabVIEW). Most of the newer spectrum analyzers and PXI instruments share a relatively similar architecture, but you need to note a few key differences. Like PXI, most traditional instruments are PC-based except that automating measurements requires the extra step of remotely transferring the result to a controller PC. Also, one unique feature of PXI is that the user can select the processor, which has the biggest impact on measurement speed. Thus, you can often upgrade PXI system measurement speed by simply replacing the controller. And several of the latest-generation PXI RF instruments feature signal processing in a user-programmable FPGA, so you can further optimize fixed- and floating-point algorithms between an FPGA and the CPU.

Figure 2. Newer-generation PXI RF instruments feature user-programmable FPGAs, high-throughput data buses, the latest microprocessors, and system design software to manage all of these elements.

These additional user-programmable (or at the very least user-accessible) components, including the CPU, FPGA, data bus, and signal processing algorithms, form the core of a redefined RF and microwave instrument architecture.

Back to Top

2. Crucial Elements of a Redefined RF System

Modular Microprocessor

Much of the time associated with RF measurements is consumed in CPU clock cycles. These calculations are math intensive, so a high-performance processor is important for making measurements such as error vector magnitude (EVM) and adjacent channel leakage ratio (ACLR). For this reason, a modular approach to not only the measurement equipment itself but also the computational engine (PC) can keep your test system performing at state-of-the-art levels for much longer than a fixed-functionality box with a CPU dating back to your original purchase. This is essentially leveraging the power of Moore’s law for RF measurements in the same way consumers have done for decades for other processing tasks on PCs.

Figure 3. By choosing a modular hardware architecture, you can upgrade the processing element of the system over time to take advantage of the latest PC technologies and achieve faster test times.

User-Programmable FPGAs

Many RF algorithms are well suited for math performed inside the data stream of the instrument. Digital upconversion, digital downconversion, and fractional resampling are just a few examples for which FPGAs have become essential to achieve cost-effective test times. Because a test system’s signal processing needs are likely to evolve over time with new standards, improved algorithms, or changes to DUT-specific schemes, the ability to modify the code inside the instrument’s FPGA can empower a test system operator to keep pace without scrapping hardware. Access to the FPGA also is crucial in transceiver systems for which decision making may need to take place quickly between the receipt of information and the generation of a response with very low latency. Testing devices such as RFID tags can significantly burden these closed-loop response times, but access to FPGAs in both the data stream and the core instrument architecture itself gives test engineers a viable approach to address them.

This is also true with “protocol aware” systems, for which asynchronous serial protocol communication is required for test. By building the specifics of the protocol into the FPGA, you can abstract the programming to a high enough level to focus on the information being transmitted or received instead of concentrating on the low-level details of the protocol communication. This can make measurement systems much more code-modular as protocols change or evolve.

Simply providing the ability to program the FPGA, however, is not enough for this capability to be useful. The programming language must be accessible and productive enough for RF test engineers, which has often been a barrier when limited to VHDL or Verilog approaches. By bridging this significant gap, the LabVIEW FPGA Module toolchain has proven crucial to advancing the implementation of user-programmable FPGAs in RF systems.

Figure 4. FPGAs have seen a similarly impressive increase in computational power (measured in GMACs) over the last decade compared to CPUs. You can use them to optimize RF measurements as well.

PCI Express

RF and microwave measurements usually involve data sets in the hundreds of megabytes, gigabytes, or even terabytes for spectrum monitoring. Therefore, you must use the highest bandwidth, lowest latency PC bus technology available to retain as much data in a continuous stream as possible. PCI Express and its evolutions through first-, second-, and third-generation transfer rates offer instrumentation as a means of moving those data sets quickly and efficiently to memory on the PC or to disk through RAID implementations for postprocessing. As data buses like PCI Express replace the “result” buses like GPIB or LAN used in traditional instruments, you can access more information and, therefore, acquire more insight into the performance of the DUT. The PXI platform has been at the forefront of integrating PCI Express with high-performance RF measurements.

Figure 5. PCI Express represents the best combination of high bandwidth and low latency to address the needs of the large data sets common in RF applications.

LabVIEW System Design Software

LabVIEW helps redefine RF test systems in at least two ways. First, the high-productivity dataflow language native to LabVIEW is optimized for RF streaming needs. Its ability to inherently represent parallelism helps the language effectively target both microprocessors and FPGAs in a way that other languages cannot. Second, LabVIEW is a system design tool that helps you manage and integrate the many microprocessors, FPGAs, I/O points, and IP pieces necessary to build a complete RF system. Often the “glue” between these elements is as difficult to develop as the elements themselves, and the ability of LabVIEW to perform that function behind the scenes can save hundreds or thousands of man-hours of code development. This productivity gain is crucial to meet the time-to-market and flexibility requirements of the modern RF measurement system.

Figure 6. The LabVIEW system design environment features multiple models of computation to target different types of processing elements.

Multicore CPUs, user-programmable FPGAs, PCI Express, and LabVIEW system design software are crucial building blocks to a redefined RF test system. National Instruments is working to empower engineers from a variety of backgrounds to use these elements with each other and with the analog, digital, and RF PXI hardware components needed for a measurement. The following examples demonstrate how you can deploy them in practice.

Back to Top

3. Examples of the Innovation of Redefined RF Instrumentation

MIMO Measurements

Multiple input, multiple output (MIMO) transmission schemes are a common way to boost bandwidth inside the same frequency range and protocol. A key requirement in this scheme is phase-coherent generation and acquisition. The traditional box instrument approach of phase-locked looping multiple oscillators to a common 10 MHz clock results in each instrument’s phase-locked loop (PLL) having different phase noise on each channel. The modular hardware approach overcomes this problem with a local oscillator shared between multiple generation or acquisition modules, thereby offering identical phase noise for the various RF ports. Figure 7 represents the difference in phase noise between using a modular architecture with a shared local oscillator and simply sharing the reference clock between several boxes.

Figure 7. Modular instruments can often share a common local oscillator as opposed to the traditional approach of phase-lock looping to a common 10 MHz clock.

Figure 8. This shared local oscillator approach offers significantly better phase offsets for high-frequency systems.

When programming MIMO systems, you should use high-level application programming interfaces and well-designed synchronization programming schemes to minimize the difficulty associated with data streams coming from so many sources. The LabVIEW environment’s native parallelism provides an intuitive way to represent MIMO systems and overcome these complexities. Both the IVI-compliant driver models of the NI-RFSA and NI-RFSG instrument drivers as well as the NI-RIO driver model for user-programmable FPGAs assist in abstracting the low-level details necessary to make these systems successful.

Mixed-Signal Automated Test (Power Amplifier Characterization)

To test a modern RF device such as a power amplifier (PA), the RF instruments (both generators and analyzers) need to work with a variety of other devices including digital generators, arbitrary waveform generators (AWGs), and battery simulators.

Figure 9. Testing a power amplifier requires a mixed-signal instrumentation system.

You need to be able to synchronize all of these instruments to optimize the test times of these complicated devices across the many different wireless protocols that the PA may support. A typical test sequence includes the following steps:


Figure 10. This flowchart represents a typical power amplifier test sequence.

As shown in the flowchart, the AWG is used for gain control management input of the bursted signal; the vector signal generator (VSG) generates multiband stimuli to represent each of the wireless bands the PA supports; the vector signal analyzer (VSA) captures magnitude and phase for power, efficiency, and modulation analysis; the battery simulator represents a power source for the PA similar to a battery (short transient times with high currents); and the digital generator controls the DUT over common digital protocols like SPI and I2C.

A redefined RF system offers an optimized approach to this multifaceted test system by varying the location of algorithms from user-programmable FPGAs on each instrument, supporting multicore processing on CPUs, incorporating PCI Express to transmit long waveforms to generators or receive large data sets from analyzers, and incorporating LabVIEW to orchestrate all of the code modules across multiple application programming interfaces to maximize code readability and support instrument upgrades over time. Using this collection of instruments inside a single PXI chassis, customers such as Triquint have dramatically reduced their PA characterization times (see Table 1).


Traditional Bench Test Time (s)

PXI Test Time (s)


GSM Test












Table 1. This table illustrates the speed improvement Triquint achieved in its PA characterization system using modular hardware and LabVIEW software. For more information, view the case study.

Back to Top

4. Instrumentation that Stands the Test of Time

The business and technology challenges associated with RF and microwave measurements in automated test environments are significant if traditional box instruments are the only tools you use. The redefined RF approach proposed by National Instruments uses open software and modular hardware with four key elements (multicore CPUs, user-programmable FPGAs, PCI Express, and LabVIEW system design software) to address the most demanding challenges in this space. As the wireless world continues to adapt to newer standards, more complex protocols, and higher bandwidths, these tools form the foundation of a test approach that can stand the test of time.


©2012 National Instruments. All rights reserved. CompactRIO, LabVIEW, National Instruments, NI, and are trademarks of National Instruments. Other product and company names listed are trademarks or trade names of their respective companies.

Back to Top

Bookmark & Share


Rate this document

Answered Your Question?
Yes No