Scalable, Communication-Constrained Control

As networked systems grow and operating conditions change, centralized coordination becomes impractical. This project advances system level synthesis (SLS), a framework that reformulates controller design in terms of closed-loop responses, enabling convex and scalable synthesis under locality, communication, and delay constraints that are otherwise difficult to handle.

We develop state-space and infinite-horizon SLS theory, removing the finite-horizon approximations that limited earlier SLS-based designs. We also extend SLS to nonlinear systems with saturation, sparse actuation, disturbance localization, and communication delays, and study joint control-communication synthesis where the communication protocol is optimized together with the controller for online, adaptive operation.

This system-level perspective enables distributed controllers that respect implementation limits while providing provable performance and stability guarantees. In recent work on safe multi-agent systems, joint synthesis reduces inter-agent messaging by 48-90% compared with benchmark methods in many settings of interest.

We are currently working with the Brazilian Synchrotron Light Laboratory (LNLS) on implementing SLS for orbit beam control in particle accelerators, with a focus on systematically mitigating data transmission delays among subsystems.

Selected outcomes. State-space theory of SLS for distributed H₂ control (Yu et al., 2021) and its infinite-horizon extension (Kjellqvist & Yu, 2022); nonlinear SLS for systems with saturation, locality, and delay (Yu & Ho, 2020); and joint synthesis of control and communication for safe multi-agent systems with minimal messaging (Yang et al., 2025). The rapid development of SLS theory led to a full-day workshop at CDC 2022, which I co-organized.

Timeline. 2020 – present.

References

2025

Safe Control of Multi-Agent Systems with Minimal Communication

M. Yang, J. Yu, and N. Ozay

In IEEE Conference on Decision and Control (CDC). 2025

PAPER BIB CODE

@inproceedings{yang2025safe,
  title = {Safe Control of Multi-Agent Systems with Minimal Communication},
  author = {Yang, M. and Yu, J. and Ozay, N.},
  booktitle = {IEEE Conference on Decision and Control (CDC)},
  year = {2025},
}

2022

On Infinite-horizon System Level Synthesis

O. Kjellqvist and J. Yu

In IEEE Conference on Decision and Control (CDC). 2022

PAPER BIB CODE

@inproceedings{kjellqvist2022sls,
  title = {On Infinite-horizon System Level Synthesis},
  author = {Kjellqvist, O. and Yu, J.},
  booktitle = {IEEE Conference on Decision and Control (CDC)},
  year = {2022},
}

2021

Localized and Distributed H2 State Feedback Control

J. Yu, Y. Wang, and J. Anderson

In American Control Conference (ACC). 2021

PAPER BIB CODE

@inproceedings{yu2021localized,
  title = {Localized and Distributed {H2} State Feedback Control},
  author = {Yu, J. and Wang, Y. and Anderson, J.},
  booktitle = {American Control Conference (ACC)},
  year = {2021},
}

2020

Achieving Performance and Safety in Large Scale Systems with Saturation Using a Nonlinear System Level Synthesis Approach

J. Yu and D. Ho

In American Control Conference (ACC). 2020

PAPER BIB

@inproceedings{yu2020saturation,
  title = {Achieving Performance and Safety in Large Scale Systems with Saturation Using a Nonlinear System Level Synthesis Approach},
  author = {Yu, J. and Ho, D.},
  booktitle = {American Control Conference (ACC)},
  year = {2020},
}