LAWRENCE – Look closely at any mobile service provider’s national 5G coverage map, and you’ll notice that large swathes of the country – rural areas – have no 5G service.
A new three-year, $1 million grant from the National Science Foundation will support the work of a University of Kansas researcher to improve the design of 5G connectivity and computing for rural areas – communities having unique network demands based on agricultural and community models of living and working.
It is one of a trio of new NSF-funded projects that will be led by Taejoon Kim, an assistant professor of electrical engineering and computer science and a researcher at KU’s Institute of Information Science (I2S).
“The 5G network system was developed by for-profit companies,” Kim said. “Because of this economic incentive, all these 5G networks have been rolled out in urban areas, but people living in rural areas still have to pay a subscription for their cell phones, including 5G service. But they haven’t benefited as much as people in urban areas – from the dissemination of information, faster speeds, the ability to transfer large amounts of data that really transform our lives on a different level.
“How can a company solve the problem in a way that allows it to generate more revenue?”
The KU researcher said a major barrier to 5G deployment in rural communities was “non-uniformity” in the spatial distribution of people as well as the demand for data on the network over time.
“In a city, the population is basically evenly distributed, but in rural areas you have one population group there, another group there – that’s spatial non-uniformity,” Kim said. .
Then there are the needs of agriculture. Automated machines, like GPS-controlled combines, perform high-level calculations and require a lot of communication.
“A huge amount of data has to go to the cell tower and then to the core network,” Kim said.
“They will also want to collect all this data to get statistics,” he said. “But this intensive use of data only occurs during the harvest period. It is temporal non-uniformity.
Kim and his team – KU EECS assistant professor Morteza Hashemi and collaborators from Purdue University – plan to streamline the design of rural connectivity and computing, partnering with California-based commercial firm Blue Danube to perform tests on a massive multi-input multi-output (MIMO) — an advanced antenna technology for wireless communications. Kim and her colleagues will use machine learning to understand how 5G can be better deployed to meet rural spatial and temporal demands.
“What is the main technology or hardware or software approach that will be different in a rural area compared to urban areas?” said Kim. “Artificial intelligence can learn this complicated and non-uniform behavior.”
The group will explore the use of AI to learn uniform and non-uniform behaviors and research approaches suitable for rural areas. Kim said one idea is to focus 5G signals more like a beam on specific communities and farms rather than providing coverage in large, mostly uninhabited areas to serve a city or two.
“Because it’s not uniform, it’s more effective if we focus the energy in a specific direction,” Kim said.
In addition to reinventing rural 5G access, two additional NSF awards to Kim will see him bolster commercial 5G networks for use by the U.S. government, military, and infrastructure operators and develop a technology for sixth-generation (6G) wireless technology.
Kim is the Principal Investigator of a one-year, $750,000 Phase One award from the NSF Convergence Accelerator program to help the Department of Defense improve end devices and 5G infrastructure augmentations, providing US military, government and infrastructure operators capabilities to operate via public 5G networks while meeting security and resiliency needs.
“The primary motivation for this project is the pursuit of ‘zero trust’ principles (an approach to information technology design requiring verification of all devices on a network) to address design weaknesses in networks. 5G. So we will integrate various security solutions to increase the trust level of 5G,” Kim said.
A third NSF grant for three years and $285,000 will support Kim’s work helping define 6G wireless communication requirements, using artificial intelligence to design advanced wireless network architecture for the microwave spectrum.
Kim said 5G can deliver orders of magnitude improvements in speed, connectivity and reduced latency. However, this improvement did not come from working in the millimeter wave, a part of the electromagnetic spectrum that has had limited research success in the United States. Rather, an increase in network throughput has come from the acquisition of new frequency bands and advances in massive MIMO technologies. .
Instead of using today’s radio access network architecture, which relies heavily on cell towers, Kim will explore how to effectively use massive multiple-input multiple-output (CFmMIMO) networks at scale. .
“We still need to consider using the neighborhood around 5 gigahertz — still microwaves, but using a different architectural network,” Kim said.
The current mobile phone network is based on all the mobile phone towers designed to serve the user within that cell.
“There is a new concept of ‘cellless MIMO’ which is to remove all cell boundaries, but we would have a very powerful CPU that controls a massive number of distributed base stations as access points,” said said Kim.
He said this work will also encompass cybersecurity aspects, building the resilience of AI algorithms and architecture, as well as cloud radio access networks.
Photo: The work of a University of Kansas researcher will improve the design of 5G connectivity and computing for rural areas, communities with unique network demands based on agricultural and community lifestyles and work. Credit: iStock.com.
Photo: Taejoon Kim is the Principal Investigator for a one-year, $750,000 award from the first phase of the NSF Convergence Acceleration Program to help the Department of Defense improve end-devices and augmentations. 5G infrastructure, providing capabilities for the U.S. military, government, and infrastructure operators to operate across public 5G networks while meeting security and resiliency needs.