Simplified Dynamic Programming for Decentralized POMDPs with Delayed Sharing Information Patterns via Change of Measure
Overview & Synopsis
In this paper, we consider decentralized discretetime stochastic dynamical optimal control problems with multiple control strategies operating under delayed-sharing information patterns, formulated within the framework of personby-person (PbP) optimality. We invoke Girsanov's theorem to characterize PbP optimality under a reference probability measure through value functions satisfying simplified dynamic programming (DP) equations, together with corresponding information states that serve as sufficient statistics for the strategies. The value functions and information states retain the fundamental properties of classical partially observable Markov decision problems (POMDPs), namely, both depend on the actions of the minimizing controls, rather than their strategies. The main distinguishing feature of our DP approach is that each control strategy estimates the unobservable state process and the private information components of all other strategies solely from its own private information and the delayed-sharing information components, using information states.
hal.science)
← / → for pages • + / - for zoom • 0 to fit width
Powered by Mozilla PDF.js v3.11
Document Information
| Document Title | Simplified Dynamic Programming for Decentralized POMDPs with Delayed Sharing Information Patterns via Change of Measure |
|---|---|
| Author / Organization | Charalambos D Charalambous, Photios A Stavrou |
| Subject & Discipline | Multidisciplinary Academic Open Research |
| Categories | Science & Mathematics |
| Publication Year | 2026 |
| License / Rights | Open Access / Creative Commons |
| Resource Verification | Listed PDF (Verification Pending) |
| Original Host Website |
HAL Open Science
(hal.science)
|
| File Format | application/pdf (Portable Document Format) |
| Language | EN |
| Catalog ID | DOC-194700 |
| Original Source Link | Visit Source Page on hal.science » |