Engineering team advancing cognitive distributed sensing with U.S. Army grant

September 13, 2023
Illustrative photo of distributed sensing scenario


A team of researchers in the UNT College of Engineering will work to further the foundational understanding of cognitive distributed sensing with support from the U.S. Army Research Laboratory.

Cognitive distributed sensing is a powerful framework for gathering, processing and analyzing data from multiple sensors in real-time. In this context, cognitive computing involves the use of advanced artificial intelligence systems that can learn, reason and make decisions, similar to human cognitive processes. Distributed sensing consists of spatially distributed sensors that can monitor and gather data from the surrounding environment.

The U.S. Army is investing a total of $13 million for the first year of the project, which will be led by Kostas Research Institute in collaboration with five universities, including UNT, Northeastern University, Northern Arizona University, University of Houston and University of Massachusetts-Amherst.

UNT’s portion will be $3.76 million this first year, with a possibility of additional funding over the next four years. Yan Huang, Regents Professor of Computer Science and Engineering, is the lead principal investigator for the research at UNT. Other UNT PIs are faculty members Heng Fan, Chenxi Qiu and Qing Yang in computer science and engineering; and Xinrong Li, Hung Luyen and Yusheng Wei in electrical engineering, as well as senior personnel Asif Baba and Tom Derryberry.

“Artificial intelligence is playing important roles in many aspects of research and scientific discoveries in our world,” Huang says. “There is a significant need to develop machine learning and AI algorithms to empower distributed sensing by enabling fast pattern recognition, efficient encoding of prior knowledge for decision-making and adaptive coordination.”

Photo of distributed sensing research team

UNT’s team will work specifically on developing foundational deep learning and real-time embedded AI-based solutions and a power-efficient mmWave hybrid beamforming system with phased array antennas for localization and swarm communication.

Localization is key in commercial, industrial and defense applications such as radio-active source tracking, emergency response and spectrum allocation, for which a group of robots or unmanned aerial vehicles needs to be able to collaboratively track a source. Enabling high throughput swarm communication without a powerful base station is important for applications such as a robotic or a UAV rescue team in infrastructure-impacted areas when large files of data must be communicated in real time between all robots and UAVs in the same network.

“A goal of this research is to develop more efficient directional communication capability in small devices, which is crucial in high mobility environments such as public safety, emergency response and many other areas,” says Xinrong Li, associate professor of electrical engineering. “Our work will help multiple agents like robots or UAVs to coordinate more seamlessly together and make decisions as one.”