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

It looks as though getting millimeter wave (mmWave) technology to work in rural areas might not be as daunting a task as it once appeared. Researchers at NYU Wireless went out into the field last summer to see for themselves, with some surprising results.

In a paper titled “Millimeter Wave Wireless Communications: New Results for Rural Connectivity,” researchers described in detail the measurements and models that are the first to validate rural millimeter wave path loss models. Ted Rappaport, NYU Tandon professor of electrical engineering and founding director of NYU Wireless, presented the research today at the Association of Computing Machinery (ACM) MobiCom “All Things Cellular” workshop in New York.

Rappaport and his team conducted a measurement campaign in Riner, Virginia, a rural town in the southwestern part of the state, using the 73 GHz mmWave frequency band. What they discovered is that for less than 1 watt of power and using 27 dBi antennas, they were able to achieve a reasonable signal strength.

“You can get over 10 kilometers, even sometimes in non-line-of-sight,” Rappaport told FierceWirelessTech. “It opened our eyes to how much range is possible when using millimeter wave in line-of-sight and non-line-of-sight conditions.”

Wireless engineers typically are concerned about rain and wind affecting millimeter wave signals. Rain can knock the signal down a lot and wind can shake narrow beam antennas, he said, but if you have enough margin built in with enough power at a base station, you can cover 5 or 10 kilometers easily even in rain. Greater antenna gain combined with commercial power levels could overcome a lot of rain and fog, he said.

“This opens up the way to fiberless backhaul using millimeter waves and gigahertz sized channels,” he said. “This work shows if you have enough transmit power and gain, you can overcome these weather events using wide swaths of millimeter wave spectrum.” It also shows that rural areas could be covered using millimeter wave spectrum at lower costs than current 3GPP models predict. Buildouts in rural areas traditionally have been a challenge due to smaller populations available to use and pay for services.

Rappaport said his team used the 73 GHz band because the work they’ve done to date in urban environments shows that 73 GHz is remarkably similar to 28 and 38 GHz in outdoor settings – 38 and 73 GHz are comparable in path loss after accounting for loss in the first meter of propagation distance.

They also figured that 73 GHz would make the test more challenging due to foliage and hills so if it were to work at 73 GHz, it would work in the lower bands. After all, the signal blockage is the most severe at the higher frequencies. The 73 GHz band is also in the FCC’s rulemaking to be considered for future use.

The FCC voted unanimously on July 14 on the historic Spectrum Frontiers plan to free up vast amounts of spectrum for unlicensed access and new 5G use cases. The move effectively quadruples the amount of radio bandwidth ever made available to the mobile industry.

What sparked this latest round of research was the channel model standard that the 3GPP approved in 3GPP TR 38.900 in Version 14. Rappaport said that model has many errors and inconsistencies in its rural macrocellular model, making it unusable above 9.1 GHz. That 3GPP model wasn’t verified with measurements, which is what the NYU Wireless researchers set out to do.

Rappaport and his colleagues are offering up a path loss model that mobile operators can use for rural deployments. The hope is that vendors and operators will see these results and use them as part of their network deployments rather than relying on what the 3GPP ratified, which Rappaport described as flawed.

For more:
- see this NYU Wireless presentation
- see the research paper

Related articles:
FCC OKs sweeping Spectrum Frontiers rules to open up nearly 11 GHz of spectrum
NYU Wireless' Rappaport encouraged by US policy on 5G spectrum
NYU Wireless sees trouble ahead for 5G unless 3GPP adopts new channel modeling methods