Strong Coordination over a Line Network

TL;DR

This paper investigates how three agents in a line network can coordinate their actions using communication and shared randomness, providing bounds on the capacity region and insights into optimizing communication efficiency.

Contribution

It introduces inner and outer bounds for the strong coordination capacity region in line networks and analyzes the role of common randomness and network topology in reducing communication rates.

Findings

Bounds on the coordination capacity region are partially optimal.

Common randomness reduces the required communication rates.

Matching network topology to behavior structure can lower inter-agent communication.

Abstract

We study the problem of strong coordination in a three-terminal line network, in which agents use common randomness and communicate over a line network to ensure that their actions follow a prescribed behavior, modeled by a target joint distribution of actions. We provide inner and outer bounds to the coordination capacity region, and show that these bounds are partially optimal. We leverage this characterization to develop insight into the interplay between communication and coordination. Specifically, we show that common randomness helps to achieve optimal communication rates between agents, and that matching the network topology to the behavior structure may reduce inter-agent communication rates.