Federated Coordination: Private and Distributed Strategy Alignment
Keywords: Coordination; Coordination strategy alignment; Privacy; Distributed
Abstract: Coordination in multi-agent systems is critical for optimizing collective outcomes and is applicable in diverse fields such as drone swarms, emergency response, and more. Despite extensive research, the distributed coordination strategy alignment problem---where all agents follow the same strategy and execute the prescribed actions without a global coordinator---remains largely unexplored, posing challenges in scalability and privacy preservation. We introduce a new research problem termed ``federated coordination", which seeks to achieve decentralized strategy alignment across distributed agents while maintaining the privacy of strategy choices. To address this problem, we propose a framework that employs an energy-based model. It facilitates decentralized strategy alignment by associating agent states with coordination strategies through local minimum energy values. We address privacy concerns through a simple yet effective communication protocol that protects strategy selections from eavesdropping and information leakage. Our extensive experimental results validate these contributions, demonstrating scalability and reduced computational demands. This enhances the practicality of coordination systems in multi-agent settings.
Primary Area: other topics in machine learning (i.e., none of the above)
Code Of Ethics: I acknowledge that I and all co-authors of this work have read and commit to adhering to the ICLR Code of Ethics.
Submission Guidelines: I certify that this submission complies with the submission instructions as described on https://iclr.cc/Conferences/2025/AuthorGuide.
Anonymous Url: I certify that there is no URL (e.g., github page) that could be used to find authors’ identity.
No Acknowledgement Section: I certify that there is no acknowledgement section in this submission for double blind review.
Submission Number: 5894
Loading