In This Story
Two College of Engineering and Computing faculty members are part of a project awarded $1.7 million by the National Telecommunications and Information Administration’s (NTIA) Public Wireless Supply Chain Innovation Fund. Kai Zeng, a professor in the electrical and computer engineering department, and Vijay K. Shah, an assistant professor in the cyber security engineering department at George Mason University, are part of a team led by researchers from Michigan State University to test Open Radio Access Network (O-RAN) components using artificial intelligence (AI).
Zeng, Shah, and their colleagues will use AI to guide the process of running hundreds of millions of test cases within complex O-RAN systems. “We can accelerate the testing process and make it automatic [with AI],” said Zeng.
The team’s testing includes three objectives: test for whether the O-RAN is secure, test the O-RAN's performance (i.e., its capacity for users in different app environments), and, finally, test for interoperability. When a system passes these myriad tests, Shah noted, “At least you can guarantee that, as long as a given environment is not very, very unique, it will work.” Supporting open and interoperable wireless networks is the aim of the NTIA’s Wireless Innovation Fund.
Inherently interoperable, O-RAN architecture is designed to comprise open and standardized interfaces for disaggregated RAN functions to lower the barrier for different vendors and third-party software developers to contribute to RAN innovation.
“It's not like the current monolithic, proprietary cellular networks,” Zeng explained. “If you purchase, say, Ericsson products, you can only use their hardware and software. But now, with the open RAN interfaces, different vendors can come into play that are just based on some generalized hardware; and they can build a lot of software applications or functions in this O-RAN architecture.” O-RAN architecture is standardized by the O-RAN Alliance, a consortium of cellular operators, large vendors, start-ups, and academic institutions.
Zeng’s 5G/NextG wireless expertise and Shah’s O-RAN research and development work in the NextG Wireless Lab@George Mason University positioned the pair perfectly to join this project. The project grew from Zeng’s working relationship with Michigan State and AT&T researchers, with whom he completed an earlier project funded jointly by the National Science Foundation (NSF) and the Department of Defense’s Securely Operating Through 5G Infrastructure program. That project developed a product called WindTexter based on generative AI. Shah’s O-RAN work began late in 2020 with seed funding from the Commonwealth Cyber Initiative xG Testbed. He went on to secure a separate NSF grant, Open AI Cellular, which resulted in the development of a novel AI-driven controller to automatically control or configure O-RAN system parameters.
If the project goes well, Zeng said, the research team would like to see their hardware and software tools guide private industry in purchasing, testing, and, ultimately, adopting O-RAN.