Hi, I'm Gongpu Chen (陈功谱).
A
I am a postdoctoral researcher at Imperial College London, working with Prof. Deniz Gündüz. I obtained my PhD degree from The Chinese University of Hong Kong in 2023, under the supervision of Prof. Soung Chang Liew. Prior to that, I obtained my Master’s degree from Southeast University in 2019 and my Bachelor’s degree from the University of Electronic Science and Technology of China in 2016.
My research interests lie at the intersection of communication, control, and machine learning. At the core of my work is a fundamental question: How does the interplay between information and control/decision-making shape the performance, efficiency, and robustness of intelligent cyber-physical systems? I explore this question from both communication and control perspectives, drawing on tools from information theory, control theory, Markov decision processes, reinforcement/machine learning, and optimization.
Research
Implicit communication in control systems
Motivated by the implicit communication observed in nature and human interactions, we developed a novel framework that
allows communication in control systems without relying on explicit communication channels.
The key idea is to intentionally deviate from the optimal control policy to encode information into the control input,
which can then be decoded by a receiver observing the system state.
This introduces a fundamental tradeoff between control and communication performance.
Control under communication constraints
Developing control and decision-making algorithms that remain robust and efficient under resource-constrained communication
is another focus of my research. This challenge is increasingly critical as the number of devices in CPS and IoT systems grows,
generating heterogeneous and unpredictable data traffic that makes it difficult to ensure reliable and deterministic transmissions for control tasks.
Communication for control/decision-making
A characteristic feature of CPS is the tight coupling between the cyber and physical spaces.
As a result of this coupling, the timeliness and usefulness of data in such systems usually
degrade neither linearly nor monotonically over time; instead, they depend intricately on the underlying physical dynamics.
Planning and learning for restless multi-armed bandits
Under construction
Publications
Pre-prints
-
GINO-Q: Learning an Asymptotically Optimal Index Policy for Restless Multi-armed Bandits
Gongpu Chen, Soung Chang Liew and Deniz Gündüz
arXiv: 2408.09882 -
Covert Adversarial Actuators in Finite MDPs
Edoardo David Santi, Gongpu Chen, Deniz Gündüz and Asaf Cohen
arXiv: 2501.17002 -
Intermittently Observable Markov Decision Processes
Gongpu Chen and Soung Chang Liew
arXiv: 2212.02752
Publications
-
Actions Speak Louder Than Words: Rate-reward Trade-off in Markov Decision Processes
Haotian Wu*, Gongpu Chen* and Deniz Gündüz
in the 13th International Conference on Learning Representations (ICLR), 2025.
[ArXiv] -
An Index Policy for Minimizing the Uncertainty-of-Information of Markov Sources
Gongpu Chen and Soung Chang Liew
in IEEE Transactions on Information Theory, vol. 70, no. 1, pp. 698-721, Jan. 2024.
[IEEE TIT] [ArXiv] -
Periodic Transmissions in Random Access Networks: Stressed Period and Delay
Gongpu Chen, Lihao Zhang and Soung Chang Liew
in IEEE Transactions on Communications, vol. 71, no. 7, pp. 4119-4132, July 2023
[IEEE TCOM] -
Uncertainty-of-Information Scheduling: A Restless Multi-armed Bandit Framework
Gongpu Chen, Soung Chang Liew and Yulin Shao
in IEEE Transactions on Information Theory, vol. 68, no. 9, pp. 6151-6173, Sept. 2022
[IEEE TIT] [ArXiv] -
Joint Scheduling and Channel Allocation for Kalman Filtering over Multihop WirelessHART Networks
Gongpu Chen, Xianghui Cao and Jiong Jin
in IEEE Transactions on Industrial Informatics, vol. 17, no. 5, pp. 3555-3565, May 2021
[IEEE TII]. -
Joint Scheduling and Channel Allocation for End-to-End Delay Minimization in Industrial WirelessHART Networks
Gongpu Chen, Xianghui Cao, Lu Liu, Changyin Sun and Yu Cheng
in IEEE Internet of Things Journal, vol. 6, no. 2, pp. 2829-2842, April 2019
[IEEE IoTJ] -
A Consensus-State-Aware Adaptive Algorithm for Tuning CSMA Parameters in Multi-Agent Networks
Mengdan Lei, Xianghui Cao, Gongpu Chen and Changyin Sun
in 2019 IEEE 28th International Symposium on Industrial Electronics (ISIE), Vancouver
[IEEE ISIE] -
Channel List Selection Based on Quality Prediction in WirelessHART Networks
Gongpu Chen, Rui Ma, Mengdan Lei and Xianghui Cao
in Journal of Communications and Information Networks, vol. 3, no. 3, pp. 49-56, Sept. 2018
[IEEE JCIN]
