Multi-Agent Cooperation and Learning under Information Uncertainty with Zheng Chen

Modern intelligent systems increasingly rely on distributed autonomous agents that cooperatively process information to accomplish a common task. Such systems often employ distributed consensus mechanisms to achieve network-wide coordination through local interactions.

This talk focuses on distributed consensus and decentralized learning under information uncertainty arising from communication constraints and privacy requirements. We first revisit the fundamentals of average consensus and show how communication noise fundamentally alters the convergence behavior of classical consensus algorithms. By interpreting consensus as a distributed optimization problem, we introduce an anchor-based framework that mitigates accumulated communication errors and achieves exact average consensus despite persistent disturbances. This decentralized-gradient-based framework also provides a clean and unified interpretation of several existing approaches that address average consensus under noisy communication.

We then extend the discussion to decentralized learning over wireless networks, where communication efficiency becomes a primary bottleneck. We present a broadcast-aware topology design together with probabilistic communication scheduling that substantially reduces communication overhead while maintaining fast convergence. Finally, we consider differential privacy in decentralized learning and demonstrate how topology-aware correlated artificial noise improves the privacy-utility trade-off.

Overall, these studies contribute to a general framework on multi-agent cooperation and learning under information uncertainty, offering new insights into the design of scalable, resilient, communication-efficient, and privacy-preserving edge AI systems.

Zheng Chen is an Associate Professor with the Department of Electrical Engineering at Linköping University, Sweden. She received her M.Sc. and Ph.D. degrees from CentraleSupélec, Université Paris-Saclay, France, in 2013 and 2017, respectively. Her research focuses on distributed and cooperative computing, optimization, and learning over wireless networks, with an emphasis on communication efficiency, security, and resilience.

She received the 2020 IEEE Communications Society Young Author Best Paper Award. She plays an active role in the international research community through editorial and scientific service, currently serving as an Associate Editor for IEEE Transactions on Wireless Communications, IEEE Transactions on Communications, and IEEE Transactions on Green Communications and Networking, as well as Lead Guest Editor for an IEEE JSAC Special Issue. She has also held leading organizational roles at major conferences, including IEEE ICC, GLOBECOM, ICASSP, SPAWC, and the Asilomar Conference, and served as Technical Program Chair of the 2022 IEEE SPS–EURASIP Summer School on “Defining 6G.”  She is a Wallenberg Academy Fellow.