Download this article in .PDF format
This file type includes high-resolution graphics and schematics when applicable.
Manuel Uhm, Director of Marketing at Ettus Research, a National Instruments (NI) company, is responsible for all marketing activities, including strategy, roadmapping, pricing, and promotions. Manuel is also the Chair of the Board of Directors of the Wireless Innovation Forum (formerly the SDR Forum). Serving on the Board since 2003, he has assumed various roles, including Chair of the Markets Committee, Chair of the User Requirements Committee, Chief Marketing Officer, and Chief Financial Officer.
CD: What are some of the advantages offered by a software-defined-radio (SDR) platform?
MU: SDR is defined by the Wireless Innovation Forum as a “radio in which some or all of the physical-layer functions are software defined.” By this definition, SDR is a technology that applies to baseband processing, not RF. Cognitive radio (CR), on the other hand, is the common industry term that refers to a radio that can dynamically access spectrum across a wide RF bandwidth to improve transmit and receive performance by avoiding interference, as well as avoiding prohibited frequencies in that location. Most CRs use SDR technology to do the baseband processing after the RF front end.
SDR and CR technology provide a number of theoretical benefits, but two in particular have proven to be the most economically beneficial:
• The flexibility to use a common radio design or architecture to address multiple market segments, thus achieving better economics of scale and increasing the probability of the product being a success.
• The ability to reuse software when porting from one SDR design to another. The savings in development cost has actually been one of the most tangible benefits realized by radio manufacturers, which is why most radios today are, in fact, SDRs, even though many of them don’t market or explicitly identify their radios as SDRs. Most of the benefit accrues in engineering productivity, lower development cost, and faster time-to-market.
CD: What advances in RF technology have most benefitted SDRs in recent years?
MU: Wideband RF integrated circuits (RFICs), such as those from Analog Devices or Texas Instruments, have resulted in SDRs and CRs that are easier to design, smaller form factor, and cheaper to manufacture. This has enabled the proliferation of SDRs and CRs. A good example of this is the increase in COTS radios based on an RFIC+FPGA architecture, such as the credit-card-sized B200mini SDR from Ettus Research. Previously, a radio with such functionality would have required multiple discrete RF components, resulting in a larger, more expensive radio. At the same time, there is still a place for radios based on discrete RF components for applications that require greater bandwidth or better sensitivity than can be provided by current RFICs.
CD: How does current SDR performance compare with performance from 5-10 years ago?
MU: Both SDRs and CRs have improved tremendously in size, power, processing performance, and cost. This is the main reason why SDR is now the de facto industry standard for baseband processing (including commercial wireless, military, and industrial applications), and CR is gaining significant momentum on the RF side.
CD: What applications are utilizing SDR technology?
MU: At this point, SDR is the dominant industry standard. Everything from wearables to cellphones to base stations to microwave radios to test equipment use SDR technology in the modem chips. CR, on the other hand, is an emerging technology. For cost reasons, the highest volume SDR applications, such as wearables and cellphones, use band-specific RF front ends, rather than a wideband cognitive radio that can scan the spectrum and choose the most appropriate band for transmit and receive. However, as spectrum becomes more of a shared resource (i.e., the Citizen’s Broadband Radio Service at 3.55 GHz in the U.S.), CR will become increasingly prevalent—even in consumer devices.
CD: Are there any examples of applications that SDRs have recently begun to exploit?
MU: SDR is literally everywhere. CR, on the other hand, is following a similar path as SDR in terms of market adoption. Today, CR is common in military applications, such as military radios, signals intelligence, surveillance, and electronic warfare (EW), which need maximum spectrum coverage and flexibility. Products from Ettus Research have been used for such applications. It is also common in some low-volume commercial markets, such as test and measurement and certain wireless infrastructure. For example, much of NI’s test equipment is both SDRs and CRs.
Cognitive radios from Ettus Research have even been used for wildlife tracking. However, it has not really penetrated high-volume commercial markets, as the cost has outweighed the potential benefits. This should change in the future, though, as 5G technologies and spectrum sharing become more prevalent, with CR being a key enabling technology for both. This will help to create the economies of scale necessary to drive the price down to a commercially acceptable point, so that the technology can bridge the gap between infrastructure and terminals/user equipment.
CD: What role do you see SDRs playing in regard to the Internet of Things (IoT)?
MU: Similar to my earlier statements, SDR is already a dominant technology for IoT infrastructure and devices. As an example, most application processors used for IoT devices use modems that are at least partially software defined. The role of CR is less certain, though. The highest-volume sensors and devices are likely to be too cost sensitive for CR for the foreseeable future. From a test perspective, however, it makes sense for the test equipment to be CRs and SDRs so that it can be used to test multiple devices using different protocols in multiple RF bands.
CD: What performance capabilities can we expect to see from SDR technology in the future?
MU: As processors continue to increase in capability and performance, SDRs will continue to reap the benefits in terms of flexibility (i.e., supporting more air-interface standards), lower power, and smaller form factors. As for CR, improvements in RFICs will enable CRs to support wider bandwidths (resulting in more data throughput) and cover more spectrum, also at lower power and smaller form factors.