Software Defined Radio Use Case for Test & Measurement

Jan 26, 2022

Software Defined Radios are becoming a valuable resource for Test & Measurement (T&M) of RF device function and capability development. These radio systems are particularly enticing due to their ability to be upgraded via software and/or field-programmable gate array (FPGA) IP cores, and thereby providing T&M engineers with extended flexibility, capabilities, and cost-efficiency.

The Current State of T&M

As anyone working in R&D at electronics companies can attest, significant delays and financial setbacks can arise if a company invests a great deal in device design, only to discover on a test bench that excessive noise, spurious signals, crosstalk, and/or power inefficiency are hindering performance. Avoiding such costly redesign time is crucial for getting products to market. As R&D is a huge expenditure for most company’s, T&M is becoming crucial to ensure development goes according to plan, particularly as electronics are working at higher and higher frequencies while meeting SWaP (size, weight, and power) requirements.

A big issue facing test engineers is that test instrumentation is not updated as rapidly as the devices and their associated communication protocols, frequency tuning, bandwidth, and data throughput needs. Moreover, electronics are required to be within tighter tolerances, meet a growing number of quality metrics, and overall, are becoming more complex devices, especially concerning RF communication capabilities.

T&M Needed in RF Engineering

Often the development of products relying on RF technology consists of proof-of-concept, pre-simulation, simulation, and then prototyping/validation (and eventually commercialization).

  • Proof-of-concept: this involves coming up with the ideas or problems to solve and the hardware/software requirements and specifications to achieve this.
  • Pre-simulation: once hardware selection and system architecture are complete, design teams need to make trade-offs between different aspects of the product, such as power versus performance versus component size. Often measurements on individual components are necessary.
  • Simulation: Once trade-off decisions are made, and the schematic and layout design are complete, simulation and virtual testing begin. For instance, this could be simulating the RF modulation schemes that will be used in the device.
  • Prototyping/Validation: once prototype boards are ready and bring-up testing is complete, testing for performance and specifications occurs. This also provides useful data for further development, or optimizing design and improving signal efficiency, the density of electronics, or developing ways to automate testing procedures.

Often simulation testing will depend on the intended application of the device. If it’s for satellite, military, or an autonomous vehicle, different RF environments will need to be simulated, such as noise sources, interference sources, or testing to see how the device behaves in the presence of various other RF protocols.

Figure 1: Example of a test bench and anechoic chamber for antenna testing (source)

Test benches often contain high-frequency signal generators, oscilloscopes, signal and network analyzers, specialized cables (SMA, BNA, etc), power meters and power sensors, as well RF couplers, circulators, and a wide range of other peripheral equipment. Sometimes equipment is found in isolated labs and RF shielded rooms as well or even anechoic chambers. These days, a wide range of automated test equipment and programming is also used.

The various types of T&M measures, equipment, and capabilities required in product development include the following:

  • Microwave Circuit Design: S-parameters, voltage standing wave ratio (VSWR), return loss, etc.
  • RF System Simulation: generating signals, modulation/demodulation schemes, waveforms, etc., that will be embedded in the product.
  • Electromagnetic (EM) Analysis: spectrum analyzers, signal analyzers, VNAs, etc. that, for instance, are required for electromagnetic compatibility (EMC) testing.
  • Antenna Measurements: near field/far-field measurements, SWR, etc. for antenna placement on the device.
  • Testing for Errors: bit error rate (BER) testing, introducing noise into signals, etc.

Challenges for RF engineers performing various testing include capturing large instantaneous bandwidths at high frequencies, streaming large amounts of RF data for further analysis, and the ever-evolving RF standards and protocols. Moreover, designing test processes that respect the sensitivity of nonlinear RF components fundamental to the operation of the device, is becoming more critical. Respecting thresholds such as gain compression points in amplifiers or mixer non-linearity is crucial in T&M.

SDRs for T&M

As we can see, T&M is an integral part of electronics design to ensure development goes smoothly. While standalone test equipment is often designed for great accuracy and specifically for the tasks at hand, new SDR-based platforms are becoming an important tool to perform and support various T&M.

An SDR contains a radio front end (RFE) and digital backend. The RFE contains the receive (Rx) and transmit (Tx) functionality over a very wide tuning range. The highest performance SDRs contain 3 GHz of instantaneous bandwidth using multiple independent channels and DACs/ADCs. An FPGA with onboard DSP capabilities for modulation, demodulation, upconverting, downconverting, and more are found on the digital boards. Furthermore, the FPGA is a configurable fabric containing logic blocks for customizations such as updating to the latest radio protocols or DSP algorithms, that are needed for the development of a new device.

Integrating SDRs for T&M is straightforward, and often reduces the amount of equipment needed since functionalities can be programmed in software. The benefits of using SDRs in T&M are numerous. Using an SDR does not lock a platform into a specific set of communications and functions. As well, SDRs can be paired with other SDRs or devices for different testing scenarios. The radio is also not limited to one transmission scheme or waveform but can be reconfigured to support new waveforms or operate as a different type of radio altogether. For instance, using different protocols on different carrier frequencies to simulate and test a device that is using GPS and LTE signals in the same device. Moreover, test data that is acquired in remote locations can be stored on the SDR itself and can store a far greater amount of data than traditional equipment.

In terms of customizations, SDRs can be developed that meet specific SWaP requirements, DSP capabilities as well as data link interface types. Due to the modular board designs, different form factors can be created to meet your testing situation, or components such as amplifiers that transmit at higher power can be embedded in the system. In terms of data links, high-performance qSFP+ network interfaces can stream large amounts of data to host systems and can be tailored for different streaming rates (10 Gbps up to 100 Gbps). As well, SDR systems can be specifically designed for ultra low latency applications.

Powerful SDRs also benefit from being UHD and GNU Radio compatible right out of the box. Various testing capabilities exist in GNU Radio such as:

  • Frequency spectrum plot: able to analyze your received signals on an amplitude (dB) vs frequency (Hz) plot for measurements of dynamic range or spurious-free dynamic range (SFDR), for example.
  • Spectrum/waterfall plot: this is essentially a spectrogram, which measures amplitude vs frequency vs time, where the color of the trace indicates the amplitude of the signal. Such plots are useful for testing signals input into a signal processor or filter and visualizing its performance.
  • Constellation Diagrams/Plots: this allows for plotting IQ signals, where the in-phase signal is the horizontal axis and the quadrature signal is the vertical axis, and dots on the diagram are samples of the signal. Such plots are important for measuring the IQ phase and amplitude imbalance of modulated signals.
  • Scope plots: this is essentially an oscilloscope, which measures amplitude vs time and can be used for measuring signal/waveform characteristics such as amplitude, frequency, rise time, a time interval of pulses, and distortion.
  • Waveform generators: allows for choosing different types of signals and waveforms, such as sine waves or square waves, or adding Gaussian white noise to a signal.
  • Other capabilities and tools include peak detectors, limiters, FFTs, filter design, math operators, modulators/demodulators, packet/framing operators, and error coding/decoding.

Figure 2: a screenshot of GNU Radio flow block diagram and testing constellation diagrams (source)

SDRs also allow engineers to perform automated testing in labs and the field. In the lab, tests can be programmed to respect power or frequency thresholds and complete tasks in a specified order such that damage to equipment does not occur. Automated testing also ensures reliability and consistency by removing the potential for human error.

In the field, SDRs can be pre-programmed to transmit signals at certain times for instance, or adding types of errors and/or noise into the signal when assessing signal robustness, for example. In addition, SDRs can be integrated into lab servers and can be accessed over IP connections, which is important for applications in remote areas or shared between users in a large lab or between different organizations.

Industries Using SDR for T&M

For autonomous vehicle testing, SDRs can test the increasing complexity of embedded software and simulate the rising number of scenarios that autonomous vehicles will encounter in the real world. This is particularly challenging due to balancing extreme-reliability demands with the low latency requirements. Testing everything from wireless security threats to sensor systems (such as LiDAR) to vehicle-to-vehicle (V2X) infrastructure standards. Autonomous vehicles require extreme safety compliance testing, to ensure the proper functioning of autonomous stopping, collision avoidance, and remaining in the correct lane, to name a few. SDRs can do everything from emulating channels to testing new modulation schemes.

For mobile phone providers, and particularly as the latest 5G networks are being rolled out, the use of SDRs for base station analysis or prototyping of 5G radio communications is critical. For example, using an SDR for emulation of a fading channel model is often done to assess path losses. Such emulators can be generated using blocks found in GNU Radio. Emulators are much more economical for companies as they are time efficient and reduce costs since they limit the amount of field testing needing to be done. Of course, when actual cell tower base stations are about to be deployed, testing of everything from site validation to coverage footprint to transmit/receive path information and bit rate errors are required for meeting the capacity and quality of service (QoS) of a mobile radio network provider. SDRs can provide such capabilities.

Figure 3: A wireless channel fading model can be created by generating a source (the sum of I and Q are the complex envelope) that follows the Rayleigh distribution. A similar model is used in the GNU Radio Fading Model block. (source)

Conclusion

As discussed, T&M is very important in the development of products and services relying on RF technologies. SDRs are an excellent tool for such T&M tasks and provide the ability to do testing in the field and labs using GNU Radio or acquiring data for further analysis. Many industries are already adopting SDRs for T&M due to their flexibility, reliability, and significant capabilities, as well as the cost savings that ensue.

This is the fourth article in a series covering SDR applications. The previous article was on the Software Defined Radio Use Case for Spectrum Monitoring.

Per Vices has extensive experience in designing, developing, and building SDRs for T&M applications in military & government, academic, and industrial applications. Click here to learn more about Per Vices.