NSF-AoF: Distributed Multi-Modal Sensing Aided Large-Scale MIMO Communications

Enabling innovative applications in autonomous transportation, industrial Internet of Things, and remote surgery imposes new data rate, reliability, energy efficiency, and latency requirements on future wireless communication networks. Therefore, the development of the next generation of wireless networks is a priority for the United States, Finland, and other countries around the world. To meet these requirements, this project harnesses machine learning and data collected from distributed sensing networks to design cutting-edge wireless communication networks. These networks will ensure consistently high data rates and reliable connections over large areas, providing excellent service to communication users. The proposed theories, algorithms, and proof-of-concept prototypes are expected to have impacts in multiple areas. For education, the project creates new course materials for undergraduate and graduate students enriching the wireless communications curriculum and offering hands-on training opportunities to build hardware proof-of-concept prototypes. Additionally, the project is expected to significantly impact technology transfer to industry in both the US and Finland, and facilitate research and development in machine learning based wireless communication networks by making all developed datasets publicly available to the wireless communication community. This collaboration between Arizona State University, USA, and the University of Oulu, Finland, holds promise for substantial benefits for both societies and the broader international community. The primary goal of this project is to enable scalable and reliable large-scale MIMO and high-frequency communication networks using distributed multi-modal sensing information. Toward this goal, this project seeks to develop a novel mathematical framework that balances the communication overhead of distributed sensing data with the benefits for wireless communication tasks. It also designs efficient techniques for extracting, compressing, and merging data across distributed sensing networks. Furthermore, the project develops a new scalable network architecture for distributed sensing-aided communications that optimizes interactions across various network components. It also devises training and learning strategies for large-scale distributed sensing and communication networks that operate within practical constraints of power, complexity, and data availability. Additionally, the project builds the world's first research platform for investigating distributed multi-modal sensing and communication networks using real-world datasets, following the DeepSense 6G framework. These research objectives promise significant advancements in wireless communication, with potential societal and technological benefits that extend globally. This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.