The future of 5G depends on software-defined radio

As it has in so many other applications before it, software-defined radio (SDR) is playing a major role in the development of 5G. In fact, without SDR the promises of 5G might not be achievable at all.

It checks all the boxes that define an affordable, efficient approach to receiver design and construction, including significant size reduction, analog and digital integration, low power consumption, the ability to use a single platform to cover multiple product lines, and reconfiguration – without replacing any hardware — via software even after being placed in service.

So, how does SDR achieve all this? Let’s dig into the details.

Before technology advances made SDR possible, the superheterodyne architecture and its variations were the mainstays of receiver design. This was the case since it was conceived in 1918 by Edwin Armstrong, father of frequency modulation (FM). To be fair, it may also be attributed to French engineer Lucien Lèvy; the debate continues.

Figure 1: The classic superheterodyne receiver

To make a long story short, the “superhet” uses frequency mixing to convert a received signal to a lower, fixed, intermediate frequency (IF) that has the same characteristics as the original (Figure 1). Note that all its components are analog.

In contrast, SDR consists of hardware, software and firmware with functions implemented through software or firmware with processing performed by a field-programmable gate array (FPGA), digital signal processor (DSP), a general-purpose processor or a dedicated application-specific IC (ASIC). As these functions are performed in the digital domain, wireless enhancements and capabilities can be made to an existing radio locally or over the air without requiring new or modified hardware. This would be impossible in a conventional receiver architecture. An additional benefit is that the hardware has a much longer service life, as it can be adapted through software to changes and enhancements in the network, such as new or modified protocols.

An SDR can be made very small because it eliminates many of the physically large analog components found in the receiver RF front end (RFFE). For example, it relies on a technique called direct RF sampling in which the input signal is converted from analog to digital form very near the antenna via an analog-to-digital converter (ADC). This eliminates the need for mixers and local oscillators, leaving only a low-noise amplifier (LNA) and bandpass filter, along with the ADC (see figure below). That said, when the original frequency is too high for an ADC to handle, it’s still necessary to down-convert it. But even then, it will likely be much smaller than its superheterodyne counterpart.

As all the digital components can be integrated into one or two devices, a complete receiver can be constructed at lower cost in the size of a postage stamp (Figure 2) — although more elaborate designs are larger and more expensive. In fact, Amazon sells a very basic one that plugs into a laptop via USB, which is less than one square inch and costs $26.

To create a transceiver the signal flow is the reverse and a digitized signal is sent to a digital-to-analog converter (DAC), amplified and filtered, converted to the desired frequency, and sent to the antenna for transmission. The size of the receiver/transmit module will grow somewhat but can still be remarkably small. 5G radios require a lot more than a $26 SDR can deliver, but the concept remains the same.

Figure 2: In its basic form, a software-defined radio receiver is much less complex than its superheterodyne counterpart.

5G presents non-trivial challenges

This description is basic and does not consider the formidable challenges that must be overcome for an SDR to achieve its intended performance. 5G is making this considerably more difficult because it covers a growing number of frequencies in bands between about 600 MHz to 7 GHz. It also employs higher order modulation techniques that are extremely complex. Even with a diminutive size, the transceiver (or transceivers) must still be small enough to fit within the confines of a smartphone, small cell or repeater, and consume as little power as possible.

The digitalization of RF fits neatly with the “virtualization” taking place within 5G networks in which analog functions are being reduced to an absolute minimum through SDNs and Network Function Virtualization (NFV). This is happening for the same reasons as SDR: 5G is a technological moving target that will change in many ways in the coming years. If designers had to rely only on hardware to make those changes, it’s possible that 5G would be far too expensive or technically impossible to deploy.

Extending SDR architecture to mm-wave frequencies

Another challenge facing designers of 5G radios is extending the SDR architecture for use at millimeter-wave frequencies, as this region of the spectrum is baked into the 5G standards. Its immense available bandwidth will be necessary to provide the extremely high data rates that are not easily achievable at lower frequencies. Designing an SDR or transceiver for these frequencies is not trivial because it requires the use of semiconductor technologies that until recently haven’t been fully utilized, among other challenges.

Nevertheless, as millimeter wavelengths are extraordinarily small it makes it possible for a receiver or transceiver to be integrated with a phased-array antenna with hundreds of elements, forming a complete communication package realized in a very small footprint. This has already been achieved by at least one vendor as well as within the research facilities at manufacturers of 5G network equipment.

What remains is to make these products deliverable in large quantities at a cost commensurate with the extraordinary number of small cells that will be required to blanket the nation with 5G coverage, as the range of millimeter-wave signals is very short. A decade ago, it might’ve been considered fantasy that any of this could be achievable, and some of the more esoteric features of 5G may have to wait until 6G sometime around 2030. There’s little doubt that they will arrive, just as many other technologies once considered fantasy have now become reality.

In the coming years, the SDR will become the receiver architecture of choice in nearly all types of systems, including all of those in this article, and the ADC will be one of the most important drivers of its success.

As the sampling rate and instantaneous bandwidth of ADCs continue to increase, it will be possible to direct sample at higher frequencies. To ensure these devices are free of spurious emissions, techniques such as high dynamic range receiver (HDRR) technology will be required well into the millimeter-wave region.

Dive deeper into 5G  

Follow Avnet Silica on LinkedIn