We live in a world that is becoming more and more connected. In fact, there are 6.648 billion smartphone users in the world, which translates to more than 83 percent of the global population. Despite such a large number of mobile devices and users, connectivity can still pose problems for consumers in rural areas.
Lingjia Liu, a professor in the Bradley Department of Electrical and Computer Engineering and director of Wireless@VT, has been awarded an $800,000 grant by the National Science Foundation (NSF) to help create next generation (NextG) mobile broadband networks that increase the availability of access to users by providing seamless wireless coverage and supporting varying service requirements.
This research is part of the NSF’s Resilient and Intelligent NextG Systems program, which combines resources and support from government agencies such as the NSF, the Department of Defense, and the National Institute of Standards and Technology with major US-based telecommunications companies such as Apple , Google, IBM, Nokia, and Microsoft. The goal is to focus exclusively on NextG wireless, networking, and computing systems that may have potential impacts for the future of NextG standards.
To improve network resiliency, Liu will develop the fundamental research necessary to integrate and operate terrestrial and non-terrestrial networks, termed Ground and Air Integrated Networks (GAINs). The project will focus on the use of artificial intelligence and advanced machine learning algorithms to improve communication and computing efficiencies under this extremely dynamic environment.
Terrestrial networks, also called ground networks, include marine and submarine servers, cloud servers, fiber optic cables, ground stations, and any other connection located on the ground or in the water. These networks have provided connection for several decades and have seen improvements over the last several years; however, signals at ground level can be blocked and are vulnerable to natural disasters or artificial infrastructure damages.
Non-terrestrial networks include both aerial and space networks. The aerial network includes high-altitude platforms, such as aircrafts and airships, that are situated in the stratosphere and low-altitude platforms, such as drone swarms. Aerial networks enhance communication performance because of their flexible mobility, but their battery capabilities are limited.
Space networks, i.e. non-terrestrial, support global information exchange and act as a “last resort” for communicating in remote areas. Computational resources in the space network are limited, but because of their accessibility and ability to transfer data to other devices quickly, computation tasks can be offloaded to ground and aerial networks.
Recognizing that each of these networks have unique benefits as well as limitations, Liu’s research will focus on integrating the ground and non-terrestrial networks to create the GAINs. Developing a solution that allows each of the networks to complement each other will provide users with improved and flexible services.
The research team includes Robert Calderbank, a professor of computer science at Duke University, who brings a different perspective because of his industry background. Before joining academia, Calderbank was the vice president for research at AT&T. Yuejie Chi, a professor in the electrical and computer engineering department at Carnegie Mellon University, contributes expertise in machine learning, data science, optimization, and statistical signal processing.
A group of Virginia Tech students will be working on the project, along with students from the other universities. Over the course of the next three years, these students will provide assistance by digging into a lot of the detailed research with guidance from their academic mentors like Liu and from the affiliated industry partners.
Shadab Mahboob, a Ph.D. student working alongside Liu, is motivated by the idea of helping bring more resilient wireless networks to life.
“This integration of non-terrestrial platforms like satellites to existing terrestrial networks is entirely new. It’s something that doesn’t currently exist — even in 5G networks,” said Mahboob. “The accessibility of this kind of network coverage is going to be revolutionary. Even in times of natural disasters or at isolated remote places, Non-terrestrial networks are going to allow for communication, which is going to truly change lives.”
Liu, Calderbank, and Chi are already collaborating on the design of wireless networks using machine learning. Their approach departs from traditional methods that use mathematical models.
“Machine learning has been adopted in many fields; however, NextG systems are very different,” Liu said. “Because the non-terrestrial network is highly dynamic and heterogeneous, real-time and resilient machine learning is key in this project.”
To facilitate such features, the researchers will incorporate a new 2D modulation technique that transforms information carried from one communication source to another. This waveform is being considered in 6G technology and research due to its robustness in high-speed velocity scenarios.
Liu is most excited about the opportunity to collaborate with academic partners, industry, and government to make these NextG systems a reality. In fact, the Resilient and Intelligent NextG Systems program undergoes three rounds of review — by the NSF, industry, and the Department of Defense — to ensure that the funded projects are relevant and will have a high chance to impact NextG technology.
“Since NextG will be governed by industry and industry standards, it is critical for academia to work and collaborate with industry partners,” said Liu. “By collaborating with industry partners, we will know what kind of problems are relevant and are important for NextG users. Our hope is that the whole wireless community will benefit from this research.”
Disclaimer: AAAS and EurekAlert! are not responsible for the accuracy of news releases posted to EurekAlert! by contributing institutions or for the use of any information through the EurekAlert system.