Inventor of Massive MIMO technology didn’t always want to call it that

Tom Marzetta
Tom Marzetta developed the concept of Massive MIMO to address the limitations of the original point-to-point MIMO. (NYU Wireless)

The originator of Massive MIMO (Multiple-Input Multiple-Output) antenna technology didn’t always want to call it Massive MIMO, but inventor Thomas Marzetta also happens to be the first one to coin that term in public.

Marzetta, who recently was named winner of the 2019 Fred W. Ellersick Prize, currently is an associate director and professor at NYU Wireless in Brooklyn, New York, where he's teaching. But back when he was at Bell Labs and just started to talk about the technology almost 10 years ago, he was considering terms like “large-scale MIMO” and “massive MIMO.” He favored the term “large scale antenna systems” (LSAS) at one point after the Bell Labs Director or Research recommended avoiding the term “MIMO.” Later, after everybody else started calling it Massive MIMO, he did as well.

In an interview with FierceWirelessTech, Marzetta said it’s gratifying to hear people like Sprint CTO John Saw talk about how they’re actually making Massive MIMO work and seeing the results in the field. Saw has suggested on more than one occasion that the name “Massive MIMO” is somewhat of a misnomer because the equipment isn’t massive. In fact, the footprint it uses is smaller than the old legacy systems. As it turns out, the name is more descriptive of what the technology can do rather than the size of the antennas.

Marzetta admitted he’s guilty of inventing that term, the process of which is spelled out in this blog.

Sponsored by VoltDB

Webinar: The Hidden Inflection Point in 5G: When the Changing Definition of Real-Time Breaks Your Existing Tech Stack

Rethink your definition of real-time to match the changing reality brought to the forefront by 5G. Your users expect milliseconds, in-event decision making. Is your tech stack ready?

The whole concept was considered rather radical thinking back in the early days. Marzetta is an electrical engineer, but when he landed at Bell Labs in the mid-1990s, he was transferred to the math center, which turned out to be a good thing because he was free to pursue what many may have considered to be wild or crazy ideas. However, some academic collaborators early on grasped the significance of what he was doing, and he said they deserve credit for helping him push the concept.

While a 2010 manuscript (PDF) is considered the seminal paper on Massive MIMO, many more papers have been published since then, including this one that lays out 10 myths of Massive MIMO and one critical question. One myth is that Massive MIMO is only suitable for millimeter wave bands—and Sprint clearly has proven that’s not true. The critical question identified in the paper was whether Massive MIMO could work in FDD operation.

RELATED: Sprint turns on ‘true’ mobile 5G with Massive MIMO and ENDC

Sprint has a lot of TDD spectrum, whereas other U.S. operators are heavy on FDD spectrum. That’s enabled Sprint to do some innovative things sooner rather than later. TDD is enormously advantageous for doing Massive MIMO and it has to do with reciprocity, according to Marzetta. As he and his colleagues explain it, Massive MIMO systems exploit the reciprocity inherent in the system to estimate channel responses on the uplink, and use the acquired channel state information (CSI) for both uplink receive combining and downlink transmit.

RELATED: Nokia’s Weldon needs 10 years to invent the future

Bell Labs President Marcus Weldon, who was instrumental in supporting Marzetta’s work over many years, acknowledged that it was never a slam dunk that Massive MIMO was going to be possible in the real world. However, a lot of work occurred over the past five years to make it work at scale, and to do so in a way where the economics pencil out.

“It’s hard and costly, and it just so happens, we managed to figure all of that out,” roughly 10 years after Marzetta came up with the idea, Weldon said. Nokia is now deploying both flavors in markets around the world—TDD is easier, but it also has systems for FDD, and it works particularly well in mid-band spectrum.

“Beyond Massive MIMO”

While years of effort went into Massive MIMO, the nature of Marzetta’s current research is something he calls “beyond Massive MIMO.”

The big question is: What's next? That’s his current pursuit. “The point is Massive MIMO is going to be with us for a long time," Marzetta said. "If people can field 64-element arrays, why not 264-element arrays, or if you really want to get ambitious, why not a conformal array that covers an entire skyscraper. So, Massive MIMO and its ramifications and extensions is going to occupy a lot of people for some years to come. The question is, though, once all of that is done, what are all these legions of researchers and development engineers going to do? They need fresh ideas.”

NYU Wireless researchers already are well on their way to studying the next frontier of terahertz.

"The approach that I’m taking is to try to actually look at some physical phenomenon that are known to electromagnetic theory people but which for some reason are not known to most wireless communications theorists,” he said, referring to the phenomenon of super directivity as an example.

It all sounds like science fiction, similar to how Massive MIMO must have sounded many years ago, but as long as there are fresh ideas to explore, you can bet Marzetta is going to pursue them.

“It’s an interesting business,” he said. “I’m having a great time.”

Suggested Articles

Ericsson announced the general availability of its Standalone 5G New Radio (NR) software for low- and mid-band spectrum.

In a milestone for 5G, global wireless standards body 3GPP finalized Release 16 on July 3, its second set of specifications for 5G New Radio (NR).

OneWeb is getting back into orbit thanks to a consortium led by the U.K. government and India’s Bharti Global.