NYU Wireless founder disappointed in ITU’s choice of mmWave specs for rural areas

NYU Wireless students (NYU)
NYU Wireless took 73 GHz measurements outside a home in rural Riner, Virginia. Image: NYU Wireless

NYU Wireless founder Professor Ted Rappaport tried to get members of the ITU delegation to consider his institution’s research on how millimeter wave technologies work in rural areas before the international standards body ratified ITU-5D channel models last week, but his efforts failed to change their minds.

The ITU—the agency whose purpose is to coordinate telecommunication operations and services throughout the world—held a plenary session in Niagara Falls on June 21, where it ratified the ITU-5D millimeter channel models for 0.5-100 GHz, closely following the 3GPP TR 38.901 models. The problem is, Rappaport claims, the rural channel models were based on research done at 2 GHz decades ago in downtown Tokyo—not exactly a rural area, and the standards organization didn’t consider the recent work he and his students conducted last summer in rural Virginia.

Since publishing preliminary results last fall, Rappaport claims “members of 3GPP and ITU delegations, except a few researchers in India, have shown no interest in considering the results or conducting their own studies in rural areas.”

In an IEEE paper recently published, Rappaport, an IEEE fellow, and George R. MacCartney Jr., an NYU Tandon School of Engineering student IEEE member and Marconi Society Paul Baran Young scholar, spell out how 73 GHz measurements in rural Virginia were used to develop a new rural macrocell (RMa) path loss model that is more accurate and easier to apply for varying transmitter antenna heights than the existing 3GPP/ITU-R RMa path loss models and may be used for frequencies from 0.5 to 100 GHz.

"Our earlier work, which was largely ignored by standards bodies, shows there is sparsity at mmWave frequencies that was not seen below 6 GHz in urban environments, and this will result in substantially more capacity than predicted by the adopted 3GPP and ITU models," Rappaport said.

Rappaport said he doesn’t know why ITU members did not consider his research; he has been reaching out to dozens of colleagues who attend ITU and 3GPP for the past year, and made a last-ditch effort before the ITU meeting last week in a final attempt to get their attention. But he says once practitioners start to use some of the models adopted by ITU and 3GPP, they’ll see the problems. “I just hope it’s before hundreds of millions of dollars are spent on equipment that doesn’t work the way they think it will,” he told FierceWirelessTech, adding that in its rush to get a solution, industry tends to ignore academic work.

“I think the standards would do a whole lot better if they considered peer reviewed literature,” he said. The problem with wireless standards organizations is “virtually none of their work is ever peer reviewed. It’s never subjected to the scrutiny or insights of external reviewers, other than the few scribes in the room.”

An ITU spokesperson was not immediately available for comment. However, Georg Mayer, the 3GPP TSG CT chairman, said standards represent the compromise of a group of players who have discussed the content practically in every detail. The standard does not reflect which proposals were not included and why not, he said.

In addition, if the standard were peer-reviewed, then many of the questions, alternative solutions and so forth would surely come up again and would delay the process, potentially indefinitely.

Mayer explained that in many standards bodies, the standards process is applied in a step-by-step (3 stage) model. At least in 3GPP this is the case, he told FierceWirelessTech. This also ensures that not only the group which is writing the actual standard is looking at it before completion, but also the experts in the other related groups.

NYU's earlier research that proposed a standard for using a 1 meter close in reference was indeed adopted by both ITU and 3GPP as an "optional" model for predicting signal strength over distance and frequency in urban and indoor environments.

NYU Wireless’ industrial sponsors include some of the same companies that send delegates to participate in the standards bodies, but Rappaport said he is not generally privy to their discussions as they relate to standards.

Prevailing wisdom among many network engineers is millimeter wave spectrum is not going to work economically in rural areas—they believe signals don’t travel far and can be impeded by trees and weather. But Rappaport disagrees and points to his experiments at his rural Virginia mountain home that showed millimeter wave technology could produce reasonable signal range in both line-of-sight and non-line-of-sight conditions. The work also found that greater antenna gain combined with commercial power levels could overcome a lot of rain and fog.

RELATED: NYU Wireless research shows real potential for millimeter wave in rural areas

Rappaport acknowledged this isn’t his first rodeo. He was one of the first people who said millimeter wave could work for mobile communications, and for many years, no one believed that. But times have changed, and now wireless operators are investing heavily in higher band spectrum, which will play a role in 5G.

Verizon recently won a bidding war for Straight Path, beating out rival AT&T with a more than $3 billion bid for 39 GHz spectrum that some said was close to worthless a short time ago.

Editor's note: Story updated June 26 with 3GPP comments.