Chalmers University of Technology and Virginia Tech have received funding from the Swedish Research Council (VR) and the US National Science Foundation (NSF) to jointly develop more energy-efficient radio transmitters for future 6G networks. The three-year project will use artificial intelligence and brain-inspired computing to reduce energy consumption and improve the performance of next-generation wireless communication systems.
The project is one of several Swedish-American research collaborations funded as part of the Swedish government’s initiative on sixth-generation wireless systems, 6G. The Swedish part of the project is funded by VR with SEK 2.5 million per year for three years, while the US research team is funded by NSF through the VINES (Verticals-enabling Intelligent Network Systems) programme. The aim of the funding initiative is to strengthen research collaboration between Sweden and the United States in technologies that will shape the communication systems of the future.
The researchers will develop a new type of radio architecture for future wireless systems operating at frequencies in the so-called FR3 range above 7 GHz, a frequency range expected to become important for future 6G networks. The solution is based on radio-over-fiber technology, in which signals are transmitted optically via fiber to distributed radio transmitters.
“By moving parts of the signal processing from the radio to the fiber network, energy losses can be reduced while the system remains compatible with today’s mobile network infrastructure. The architecture also provides greater flexibility as future networks are expanded to handle ever-increasing amounts of data,” says Christian Fager, Professor of Microwave Electronics at Chalmers and one of the project leaders.
Networks that can learn and adapt
One of the largest consumers of energy in today’s mobile networks is the power amplifier, the component that amplifies the radio signal before it is transmitted from the base station. As networks move to higher frequencies, this challenge is expected to become even greater.
To address the problem, the researchers will develop new radio transmitters that combine radio technology with AI-based control.
“The goal is to develop an AI-driven radio access network, known as AI-RAN, in which wireless systems can continuously learn from their surroundings and adapt to changing conditions in real time,” explains Lars Svensson, Associate Professor of VLSI Systems at Chalmers.
The project combines advanced radio hardware with neuromorphic computing, a technology inspired by the way the human brain processes information. Together with programmable circuits known as field-programmable gate arrays (FPGAs), the technology will be used to optimise the operation of the radio transmitters in real time.
International collaboration on future communication systems
The research team at Virginia Tech consists of Professors Jeffrey Walling, Lingjia Liu and Yang Yi, all at the Department of Electrical and Computer Engineering.
In addition to developing new technology, the collaboration between Chalmers and Virginia Tech will contribute to knowledge exchange between Sweden and the United States in an area considered strategically important for the digital society of the future. The research project also includes training of doctoral, master’s and bachelor’s students in FPGA technology, circuit design and next-generation wireless systems.
Source: Chalmers University of Technology
