Bringing RF into the Embedded World: It's Time

Embedded systems have been almost entirely digital throughout their long history, while RF and microwave technologies were separate subsystems with no effective interface between the two. For many reasons, this “RF/digital divide” should finally be connected.

Mercury Systems, which makes embedded systems (i.e., board-level digital and RF subsystems), proposed that manufacturers of embedded and microwave subassemblies participate in its initiative called OpenRFM. The company's goal is to make it possible to integrate RF and microwave technology into current “digital-only” embedded form factors for the first time. If you're not in the embedded systems business, you're probably wondering, “Why is this just happening now?” If you're into the designing of RF and microwave subsystems, you might ask, “Why bother?”

The answer to the first question is that digital and microwave designers have always operated in separate domains and ignored each other. The answer to the second is that integration of these two disparate technologies simply makes sense regardless of what “camp” you reside in, and for the following reasons:

  • It would allow these two technologies (digital and microwave) to be integrated in a standard form factor followed by the embedded systems industry.

  • It would be a major step toward realizing two major goals of the U.S. Department of Defense (DoD): to dramatically increase function integration, and to allow radar, electronic warfare, and other systems to be used in multiple platforms without major redesign.

  • All embedded systems manufacturers and prime contractors could follow a single design roadmap while retaining the flexibility to differentiate their products from others by using their own OpenRFM-compatible products.

  • Systems could be constructed that are smaller, lighter, consume less power, are less expensive, and shorten the time to market for both embedded systems and end products.

Although Mercury targeted OpenRFM to sectors of the defense industry in which it participates, there is no reason why it could not be adopted in other embedded markets such as rugged industrial, scientific, and medical systems, and telecommunications, for example. As some of these applications make use of wireless communications in some form, OpenRFM should be useful for manufacturers of commercial systems that incorporate it.

To affect this integration on a broad scale, Mercury will need to have it appended to the current family of standards used in defense embedded systems — OpenVPX, whose champion is VITA (formerly the VMEbus International Trade Association). There is at least a reasonably good chance this will occur as it fits neatly into the association's roadmap, which has already expanded to include software-defined radio and space qualification.

Figure 1. This is an OpenRFM module riding atop an OpenVPX board. Traditionally, the functions performed in the OpenRFM module would have been implemented in a larger separate housing.

OpenRFM is a modular, open architecture built around OpenVPX that combines hardware, firmware, and software, and allows high channel density, advanced interconnect technology, and employs a “building block” approach. The company has already produced OpenRFM products in 3U and 6U OpenVPX form factors, as shown in Figures 1 and 2. The real estate available for OpenRFM in the OpenVPX environment is sufficient to allow a broad array of integrated microwave subsystems (today referred to as Integrated Microwave Assemblies or IMAs) to be accommodated.

One of the most common is the downconverter, which effectively forms the front end of every system in which a signal must be captured over the air and lowered in frequency to one that can be handled by an analog-to-digital converter. The downconverter's opposite is the upconverter, which does just the reverse: taking a signal at a lower frequency and increasing it to one required by a specific application. Both of these IMAs are required elements of every type of military and commercial system, from satellite communications terminals to microwave point-to-point links, and almost every type of communications system.

Figure 2. The 6U VME carrier (bottom) and three OpenRFM modules (top): a quad downconverter (left), wideband tuner (center), and direct-digital synthesizer (right).

These are complex products that include many active and passive microwave components, and range from relatively small to quite large, depending on their requirements. As these subsystems typically don't reside on a board-level product but rather are separate functional blocks, OpenRFM has the potential to make the whole system smaller. Other typical IMAs include complete digital receivers that, when combined with traditional signal processing, general-purpose processing, signal distribution, and other functional blocks, form a larger subsystem.

Why Is This Just Happening Now?

As we are all painfully aware, the defense industry has been driven primarily by technology rather than cost. This is because new radar, electronic warfare, and other systems required by the DoD are extremely complex and push the envelope of achievable performance. It's also because the Pentagon has to combat both current and perceived next-generation capabilities of its adversaries and because these systems remain in the field for decades with upgrades over the years. Both of these requirements don't come cheap.

However, the days of this “just do it” mentality are rapidly coming to an end as the DoD scrambles to squeeze as much as possible from its annual budgets. It's obviously not the first time Pentagon spending has been under the microscope, but it may be the first time that, technologically speaking, there are ways purveyors of digital and microwave technology can broadly help make it happen. OpenRFM is well suited to become one of these enablers.

By standardizing the electromechanical, software, control plane, and thermal interfaces used by IMAs, OpenRFM could almost completely change the way defense systems are built, providing a standards-based roadmap that would allow systems to be used on more than one platform, which is the opposite of why they are designed and built today. Even though a radar built by manufacturer A for one aircraft may perform very similar functions as one built by manufacturer B on another aircraft, they are invariably almost totally different, incorporating proprietary designs that can't easily be “ported over” from one platform to the next.

There Are Precedents…Sort Of

There are market sectors in which digital and microwave technologies comfortably coexist in a quasi-standard form factor — smartphones and other wireless devices being a classic example. In all these cases, integration was driven by need: There's simply no room in these small, typically battery-powered devices for disparate technologies singing different tunes.

Bluetooth products and Zigbee owe their very existence to wireless communications, as do Wi-Fi access points, some medical devices, and other products in which size or some other constraining factor dictates that digital/microwave integration is essential. Their small size and specific digital and microwave requirements allow this integration to be accomplished at the wafer level using a System on a Chip (SoC) approach that can integrate a truly astonishing amount of functionality in a single device. A good example is Silicon Labs’ Si468x CMOS digital receiver ICs that are essentially low-power, multi-band, digital broadcast receivers that support AM, longwave, shortwave, FM, AM and FM HD radio, digital audio broadcast, and digital multimedia broadcast.

Among other things, they include an OFDM channel demodulator, audio DSP processing, on-chip source decoding, I2S digital audio output, a stereo audio DAC, VCO, PLL, frequency synthesizer, AGC SPI, and I2C control interfaces. Typical applications include multimedia handsets, media players, GPS devices, tablets, all types of fixed and portable radios, boom boxes, and entertainment systems.

One other SoC that packs broad functionality in a single device is Broadcom's BCM20736 WICED SMART Bluetooth SoC that allows OEMs to develop applications on its ARM Cortex-M3 processor. It includes an RF and embedded Bluetooth stack, support for wireless charging, software for profiles, the stack, APIs, and application software development kits, and has two serial peripheral interfaces (SPIs). The device, designed to be powered by a single coin cell at 1.2 VDC, is pin-compatible with Broadcom's Bluetooth Smart SoCs, and allows secure updates to be implemented over the air.


It's important to note that other defense electronics manufacturers have, out of necessity, developed their own ways to integrate RF and microwave technology onto board-level products. The difference is that, unlike Mercury, none has gone so far as to put a standards proposal on the table.

The first major challenge is getting the member companies within VITA to give approval for incorporating OpenRFM in OpenVPX after the usual standards development process. Interestingly enough, OpenVPX was also proposed by Mercury in 2009 as a way to broaden the acceptance of the VPX standard, and a working group of 28 companies participated in its development. The challenge will be to get many of these same companies — almost none of which have interest or technical capability in RF and microwave technology — to buy into the concept. If they do, they would effectively be signing up to adhere to the standard if and when they do choose to add this technology to their product portfolio.

If OpenRFM takes root after adoption by VITA, the next and unquestionably most formidable challenge will be enticing the microwave industry to participate. This will be orders of magnitude more difficult than with OpenVPX for two reasons. First, VPX was already a VITA standard and required introduction of no new or drastically different technology. The challenge was getting Mercury's competitors on board. Second, the microwave industry, at least those parts of it serving aerospace and defense, has no standards body like VITA and is unaccustomed to dealing with any standards other than those dictated by DoD or other government agencies. It also has no relationship with VITA or the embedded systems business. Consequently, OpenRFM is likely to be a tough sell to these companies, although it may not initially matter whether this industry participates or not as it takes its orders from DoD and prime contractors; however, if DoD mandated its use, IMA manufacturers would have to conform to it.

Limitations and Opportunities

Figure 3. Silicon Labs Si468x Digital Radio Receivers are an all-in-one solution.

OpenRFM is inherently best suited for applications in which RF power output is low, as higher power requires larger components, resulting in subsystems too large to be integrated within OpenVPX. This means it will be used in receive rather than transmit sections of systems and in the low-power driver stages of RF power amplifiers, low-noise amplifiers, and other small-signal subsystems.

It would, however, be appealing for use in subsystems operating at very high frequencies (the millimeter-wave region), where components are much smaller than at lower frequencies, and power levels are lower as well. While it may not be possible to incorporate higher-power RF and microwave subsystems within Open-RFM, it may still be possible to make them compatible with it, which would still allow systems from various microwave manufacturers to build to the standard.

What's Next

OpenRFM was first exposed to the embedded community at a VITA meeting called Embedded Tech Trends. For the microwave industry, however, the next step is for Mercury to present its case for Open-RFM at IEEE's International Microwave Symposium. Sponsored by IEEE, the industry's primary international symposium and exhibition brings together almost the entire industry throughout the world. At this year's event, the company introduced its first 500-to-18 GHz digital microwave tuner targeted at electronic warfare applications. The Ensemble RFM-1RS18 consists of three OpenRFM modules in a single-width 6U VXS-format package. Within this space is a converter with four IF outputs and a frequency synthesizer.

It's important to note that Mercury wins regardless of whether or not Open-RFM is accepted within the embedded systems and microwave communities or is made part of OpenVPX, as it can still use this formula for its own products. Nevertheless, the impetus for integrating RF and microwave IMAs within boardlevel products that, with few exceptions, are totally digital, was to meet the looming challenges facing both industries as DoD is serious about changing the design paradigm. It makes this case at every available opportunity, in programs conducted by the Defense Advanced Research Projects Agency (DARPA) and at R&D laboratories in the Army, Air Force, and especially the Navy. It may take some time before the defense industry broadly accepts this challenge, but in the long term, it has little choice.

Another factor behind the development of OpenRFM was the most basic: to integrate digital and microwave technologies that have existed in their impenetrable stovepipes since World War II. After all, design engineers are only human and typically resistant to the disruptive challenges of change. Without some force driving these two technologies together in the embedded space, who knows when — or if — it would ever occur, and OpenRFM may just be such a catalyst.

This article was written by Barry Manz for Mouser Electronics, Mansfield, TX. Reprinted with permission from Mouser Electronics. For more information, visit here .