5G OTA Test Not Ready For Production

5G is poised to dominate the wireless world, but over-the-air (OTA) testing of 5G beamforming antennas is still not ready for volume production.

Beamforming is a critical element in the millimeter wave version of 5G, because of the limitations of ultra-high-frequency signals. Unlike 4G and its predecessors, millimeter wave technology will not penetrate objects, so signals need to be directed around objects. In addition, signals will not travel as far without interruption, requiring a large number of small cells for transmitting signals and an array of antennas to receive those signals. But because the leads of those antennas are not exposed—in many cases, they are embedded into the package—the only way to test them is over the air.

The problem is that test and validation providers have yet to come up with a high-volume-production-ready system for over-the-air testing of 5G systems. While most big test equipment companies support sub-6GHz 5G devices—the first phase of the 5G rollout— equipment makers are still scrambling to create test solutions that are fast enough to handle high-volume production for the second phase of 5G, millimeter wave. That’s the one that really pumps up the speed and data capacity of wireless communication, opening up new opportunities for connected devices, buildings and infrastructure.

“At first there were a lot of concerns about OTA and how to handle that,” said David Hall, chief marketer at National Instruments. “Lab-based OTA test systems has become fairly widespread, but the current methodologies used in the lab environment do not scale to the cost and speed expectations of the manufacturing floor. As a result, NI continues to investigate both near-field and far-field approaches to OTA testing in preparation for delivering OTA-based manufacturing test solutions in the future.”

Other test vendors and OEMs already have shipped or announced a host of products during the past few months that highlighted a few specific capabilities in testing mmWave 5G products. Most emphasized the range of their products that support testing of sub-6GHz 5g products, which are more familiar, simpler, and a much better business proposition for most companies. For now, the big growth in 5G continues to be in the mobile consumer market, according to Earl Lum president of telco consultancy EJL Wireless Research.

Still, the mobile industry has little choice but to push into the millimeter wave space.

“The reason that it is worth taking the challenge up is that we have nowhere else to go,” said Neill Mullinger, product manager for the vertical market solutions group at Mentor, a Siemens Business. “We’ve used up all the spectrum, but demand is still skyrocketing. If you look at any city in the world, you test out the performance of your cellular phone in 3 a.m. and it’s pretty good. You try it out at 6 p.m. and it is lousy. It’s the sheer volume of users. It’s like being on a congested highway. You can drive freely at night, but you can’t at 6 p.m. It’s the exact same problem, so we have to go there. We have to go into the millimeter wave.”

More data everywhere
Part of the problem is more data per user, and this is why 5G is so critical. While the growth of mobile users is flattening, 5G will provide a big bump in sales and capability.

Market-research firm Canalys predicts the market for 5G-enabled handsets will reach 800 million units in 2023, accounting for 51.4% of a smartphone market. Ericsson, meanwhile, predicts the number of network subscriptions will grow from 10 million by the end of this year to 1.9 billion by 2024.

The influx of new business is critical for an industry that has seen growth flatten and actually decline over the past couple years. International Data Corp. issued a report in May forecasting a 1.9% drop in smartphone shipments in 2019 compared to 2018. It will be the third consecutive year of market contraction, according to the firm.

This involves more than just handsets. “When you have all those mini cells doing this, the 5G solution is actually not just wireless,” said Peter Zhang, R&D manager in Synopsys’ Solutions Group. “If people just purely looking at the wireless portion of the 5G’s impact on the communication network, that’s actually limited to what only one aspects of that. Let’s not forget, there was also the gigantic backbone land line networking systems that we’re constructing and working. And we all have to be working together congruently with our 5G wireless portion of it so we can access a data network everywhere with that high-performance type of experience.”

That may be a couple steps down the road. Much of the initial growth will be in the sub-6GHz realm, especially in the consumer market, 5G is an extension of 4G, especially since the standard was modified to allow sub-6GHz 5G devices to function with 4G infrastructure. That makes for a much cheaper, smoother migration than anything involving mmWave, according to Ajit Paranjpe, CTO of Veeco.

Still, it won’t be easy to skip the difficulties of the mmWave market in the long run, Paranjpe said. Sub-6GHz integrates with 4G, but includes the same beamforming, massive MIMO, high-bandwidth, low-latency requirements as the higher end of the 5G spectrum.

Finding a way to test mmWave antenna/RF IC components proved to be so different from previous generations that no existing off-the-shelf test equipment was up to the challenge, a June, 2018 study published in IEEE Access.

At least part of the problem is that 5G is a system of systems. “It involves a lot of new moving parts,” said Ian Dennison, senior group director for customized IC and PCB group at Cadence. “We’ve got the optical front haul to the small radio heads, and there are many, many of them. We’ve got this concentrated radio access network baseband edge compute going on. And this whole system of systems needs testing. And it’s going to need some sort of over-the-air test. We can do the emulation and FPGA prototyping and simulation, of course. But in addition, it is going to be a lot of real-world testing. Part of this expectation-management problem is that we should expect that our 5G systems start to improve over time as the AI systems start to kick in, start to develop better quality of service. But straight out the box it may not be perfect, so testing is going to be a continuous thing—and something that the network operators are going to be relying on to improve the quality of their service.”

Spotty progress
There are plenty of customers already doing private networks, experimental projects or just moving ahead with mmWave-based networks for companies, industrial automation or fixed wireless. And there are enough other things in the works that test companies don’t have the option to put off support for mmWave. All of this requires full-scale production testing capability for both mmWave and sub-6GHZ, according to Clinton Linville, application engineer for the Systems Group at Rohde & Schwarz.

“Millimeter-wave-type products are becoming very complex because we’re dealing with these phased array antennas, which are going to have multiple elements, each of which will have some kind of phase shifter and possibly a power amplifier,” Linville said. “The phased arrays steer the signal in the direction of a phone or other node, so it’s important that we understand how they are performing. If we test using a connector to the radio and another to the antenna, we can’t really quantify how the directed beam is going to look. It might not have the proper gain or the components behind it might be too noisy and could impact the modulation of the signal. So we have to look at the radio plus the phased array antenna together because that is going to tell us the real performance of the system.”

The picture gets a lot more complicated once the market advances beyond the need to give consumers a phone just like their old 4G but faster, and starts requiring that tech providers get involved with more sophisticated capabilities that exist at both ends of the 5G spectrum.

“There are a ton of people out there doing chips focused on beamforming, phase and amplitude adjustment, because the antennas are huge on the array side and they need a lot of chips,” said EJL’s Lum. “The testing side is coming along, but it’s a lot more mature below 6GHz than for mmWave because there isn’t that much activity in millimeter wave, by comparison.”

Trying to shift from processes based on tests using physical probes and relying instead on over-the-air tests of devices with the kind of dramatic attenuation the 5G mmWave frequencies demonstrate – plus its host of other idiosyncrasies – was a challenge from the very beginning, according to Michael Foegelle, director of technology development at test-components provider ETS-Lindgren.

“The biggest challenge to our engineering teams and across the industry is simply the higher complexity. 5G brings many new bands, most of which have much wider bandwidth signals than anything that we have seen in LTE,” said Ben Thomas, director of technical marketing at Qorvo, in a recent presentation. “Combine that [complexity] with many more carrier aggregation combinations, dual uplink, and more complex wave forms and modulation schemes, and quite frankly you’re looking at an exponential impact on the RF section.”

It’s not clear what enhancements, specifically, it will take to create a high-volume-production-capable automated test system for an OTA test process, partly because the system would have to adjust to the manufacturing requirements of customers, Linville said.

A robot arm that could place a component in the very precise spot required in current, anechoic-chamber-based systems as part of a larger production line might be appropriate, for example. But it also is possible that the lessons already learned about how to test mmWave antennas could be modified for efficiency’s sake, Linville said.

“We haven’t really looked at what 5G MIMO might look like for FR2 (mmWave), because then we’d have to consider multiple angles of arrival for the signal and be able to apply the proper fading channels,” he said. “There are still some basic characteristics they’re trying to nail down for over-the-air testing. We have to look at making sure we know if a 4G LTE radio is impacting something in the 5G radio, for example. There are so many considerations that we have to we have to keep in mind because it’s you know it’s new technology. It’s a new way to be able to look at these devices as a whole because historically we’ve been able to separate these things out. We’ve got the radio side of it. We’ve got the antenna side of it. And then you can just do some basic things to marry them together. But with 5G and millimeter wave, there is so much more involved with the physics. You know the free-space law of your test environment is going to be really important.”

Learning curve
In the meantime, this has put a renewed emphasis on simulation. All of the big simulation vendors—Cadence, Synopsys, Mentor and ANSYS—have seen an uptick in simulation business related to 5G.

This includes 5G chip errors, but it also includes complex, mesh-network-based M2M connections that might not have been relevant with other technologies, said Frank Schirrmeister, senior group director for product management and marketing at Cadence.

“We see customers testing a regular framework for a flow, but also a different level of testing in simulation – like how do you test, if you suddenly have a swarm of a thousand small objects, to make sure they can talk to each other,” Schirrmeister said. “From what we’re seeing, the rollout, the way 5G is implemented, will be very application independent and rely heavily on private applications an networks that may not be so visible to others.”

The learning curve for customers and colleagues also can be an issue, according to Keith Cobler, marketing manager for WIC & MNT at Rohde-Schwarz. “You’re seeing now a big difference. In RF engineering the wavelengths are very long and basically everything has been governed by E=IR at the circuit level. But as you go to millimeter wave, everything acts like an antenna. It acts very different. It’s actually a huge learning curve. We’re basically going from We’re going from E=IR to Maxwell’s equations.”

It’s hard to say how long it will take before production-level testing is available for 5G, even from companies with plenty of experience building systems for manufacturing.

“People are already doing some advertising, but the device pool is still quite small,” Linville said. “Once 5G becomes more mainstream, then you’ll see that people will have forced the hand of the test instrument industry and we’ll have something easy to plug into a production environment.”

—Susan Rambo and Ed Sperling contributed to this report.

Related Stories

Gaps In 5G Test

https://semiengineering.com/knowledge_centers/communications-io/wireless/5g/