Smoothing the Bumpy Road to 5G Wireless

When 5G wireless technology debuts in 2020 or so, a whole new world of dazzling technical possibilities will open up. Downloads will be infinitely faster; batteries will last ten times longer; bandwidths will soar. And all this will happen while 50 billion devices get connected.

That's the good news. The not-so-good news is that the road to wireless nirvana is going to be a bumpy one. The task of designing, building, testing, and verifying 5G wireless prototypes may unfortunately turn out to be one of the most complex in recent engineering history.

That's why suppliers have begun stepping in to help. By creating new ways to develop 5G software, they're enabling technical teams to break down the walls between engineering disciplines. More important, they're speeding up the prototyping process.

"The only way to really address the challenges of 5G is to have some kind of integrated platform that allows the different disciplines to work together," noted James Kimery, director of marketing for wireless research at National Instruments. "Otherwise you're going to be re-inventing the wheel every time you want to prove a concept."

Such integrated platforms are critical because they can quell the worst fears of all 5G engineers, Kimery said. "The one thing they don't want is to spend months in prototyping, get to the end of the process, and realize their concept doesn't work."

A Classic Engineering Problem

To understand why that's a concern, it's best to begin by discussing 5G engineers' soaring ambition. They want to cut the download times for a two-hour HD movie from 30 seconds to three seconds. They want to boost data rates from 150 Mb/s to 10 Gb/s. They want to extend connectivity to 50 billion devices, many on the Industrial Internet of Things (IIoT).

Unfortunately, such lofty ambitions are inextricably linked with uncertainty. "On 1G, 2G, 3G, and 4G, there was an engineering consensus at every level, in terms of goals and objectives," Kimery said. "But on 5G, we have no consensus at all."

READ MORE ARTICLES ON WIRELESS:

Then, there's the chaotic effect on the engineering process. 5G wireless technology touches a wide variety of technical fields, necessitating levels of expertise that are both broad and deep. Experts in RF and wave propagation are obviously needed, but the list doesn't stop there. Software expertise is critical, and equally important is an in-depth knowledge of digital signal processing and field-programmable gate arrays.

The problem is that virtually no engineer can master all those fields. "Some of the technical problems can be outside an engineer's comfort zone," noted Ken Karnofsky, senior strategist for signal processing applications at The MathWorks Inc. "The DSP guys may know nothing about RF. And the RF guys may know nothing about FPGAs."

The result is a classic engineering dilemma on a grand scale. Algorithm developers throw their work over the wall to the DSP engineers, who throw it over the wall to the engineers running the FPGA tools, who complain that the algorithm doesn't fit. The problem is compounded by the fact that semiconductor companies, infrastructure providers, handset makers, and service operators are all approaching the task from different directions.

That's where integrated software suites and simulation tools come in to smooth the work flow. They allow software algorithms to be more easily incorporated into system designs, prototyped onto hardware, and moved into implementation.

National Instruments launched that concept more than a year ago, when it introduced a product called the Labview Communications Systems Design Suite, which offers an environment that's closely connected with the company's software-defined radio hardware. The fundamental advantage of LabView Communications is that it ties together the tools, which would otherwise be radically different, at each stage of the process. As a result, each step - from algorithm development to FPGA deployment to test and validation - takes place in a common, graphically-based setting.

Industry analysts consider the availability of such tools a critical step for 5G development. "The graphical canvas allows domain experts to program FPGAs themselves, eliminating the need to bring in specific programming expertise, [and] saving time and money during the design cycle," wrote Patrick Moorhead of Moor Insights and Strategy in a research paper called "The Rx for 5G RF."

Faster Prototyping

The same approach is being employed by The MathWorks, in its well-known Matlab numerical computing environment, and more recently in a product called SimRF. An add-on to the company's well-known Simulink product, SimRF provides a component library and simulation engine specifically geared to RF systems. In a graphical environment, it incorporates amplifiers, mixers, and other basic building blocks for designing wireless transceivers. More important, it provides a common setting for DSP and RF engineers.

"It's a single environment where they can work together," noted Karnofsky of The MathWorks. "The signal that's generated from one environment can go directly into the RF model, come right back out again, and provide results. It's an improvement over the way it used to be done. In the past, with separate tools, you had to make assumptions that may not have reflected reality very well."

For 5G developers, the bottom line is speed. The MathWorks says that SimRF is much faster than traditional circuit simulation. The technology is already being used in simulation of millimeter wave (mmWave) systems, which would enable wireless networks to go from 2 GHz bands today to as high as 72 GHz in 5G.

Similarly, Nokia Networks has prototyped mmWave communication links using LabView Communications. The company has developed one of the first mmWave communication links capable of streaming 10 Gb/s, reportedly making it the fastest mobile access wireless system ever demonstrated.

LabView Communications also served as an enabler for another 5G candidate technology called FD MIMO (full-dimension MIMO), which calls for a 32-fold increase in the number of antenna elements in 5G base stations. At the recent NIWeek conference in Austin, Texas, Samsung demonstrated a working prototype of FD MIMO, developed in conjunction with NI.

"It showed they were able to take a complex idea and prototype it in about a year," Kimery said of the effort. "They were happy because they expected to spend a lot more time and money than they did."

Experts say they don't know whether the new breed of tools will enable a 2020 rollout for 5G, but they're sure that engineering teams will reap benefits by using them. "It's impossible to say how much this will speed up 5G research," Karnofsky concluded. "But evidence from past generations of research suggests that this is an improvement over a piecemeal approach."

Senior technical editor Chuck Murray has been writing about technology for 31 years. He joined Design News in 1987, and has covered electronics, automation, fluid power, and autos.