The distributed dual ascent algorithm is robust to asynchrony

TL;DR

This paper introduces an asynchronous version of the distributed dual ascent algorithm that converges despite heterogeneous delays and lack of coordination among agents, enhancing robustness in multi-agent optimization.

Contribution

It presents a novel asynchronous dual ascent algorithm that works under heterogeneous delays without requiring agent coordination, unlike previous randomized schemes.

Findings

Convergence is proven under heterogeneous delays.

Algorithm operates without coordination among agents.

Robustness to communication and update frequency variations.

Abstract

The distributed dual ascent is an established algorithm to solve strongly convex multi-agent optimization problems with separable cost functions, in the presence of coupling constraints. In this paper, we study its asynchronous counterpart. Specifically, we assume that each agent only relies on the outdated information received from some neighbors. Differently from the existing randomized and dual block-coordinate schemes, we show convergence under heterogeneous delays, communication and update frequencies. Consequently, our asynchronous dual ascent algorithm can be implemented without requiring any coordination between the agents.