Abstract:Learning-to-communicate (LTC) in partially observable environments has received increasing attention in deep multi-agent reinforcement learning, where the control and communication strategies are jointly learned. Meanwhile, the impact of communication on decision-making has been extensively studied in control theory. In this paper, we seek to formalize and better understand LTC by bridging these two lines of work, through the lens of information structures (ISs). To this end, we formalize LTC in decentralized partially observable Markov decision processes (Dec-POMDPs) under the common-information-based framework from decentralized stochastic control, and classify LTC problems based on the ISs before (additional) information sharing. We first show that non-classical LTCs are computationally intractable in general, and thus focus on quasi-classical (QC) LTCs. We then propose a series of conditions for QC LTCs, under which LTC preserves the QC IS after information sharing, whereas violating them can cause computational hardness in general. Further, we develop provable planning and learning algorithms for QC LTCs, and establish quasi-polynomial time and sample complexities for several QC LTC examples that satisfy the above conditions. Along the way, we also establish new results on a relationship between (strictly) QC IS and the condition of having strategy-independent common-information-based beliefs (SI-CIBs), as well as on solving Dec-POMDPs without computationally intractable oracles but beyond those with SI-CIBs, which may be of independent interest.
| Comments: | Preliminary version appeared at IEEE CDC 2025 |
| Subjects: | Systems and Control (eess.SY); Machine Learning (cs.LG); Multiagent Systems (cs.MA); Optimization and Control (math.OC) |
| Cite as: | arXiv:2603.03664 [eess.SY] |
| (or arXiv:2603.03664v2 [eess.SY] for this version) | |
| https://doi.org/10.48550/arXiv.2603.03664 arXiv-issued DOI via DataCite |
From: Kaiqing Zhang [view email]
[v1]
Wed, 4 Mar 2026 02:36:18 UTC (1,270 KB)
[v2]
Wed, 29 Apr 2026 03:31:49 UTC (1,269 KB)
此内容由惯性聚合(RSS阅读器)自动聚合整理，仅供阅读参考。 原文来自 — 版权归原作者所有。