DIMIX: DIminishing MIXing for Sloppy Agents

TL;DR

This paper introduces a two time-scale decentralized algorithm for non-convex distributed optimization over time-varying networks that tolerates lossy, noisy, or compressed information sharing, achieving convergence despite communication constraints.

Contribution

It proposes a novel two time-scale method that handles lossy information sharing in non-convex distributed optimization over dynamic networks, with proven convergence guarantees.

Findings

Achieves convergence rate of O(T^{-1/3 + ε}) under lossy communication.

Supports noisy, quantized, and compressed information sharing methods.

Simulation results validate theoretical convergence guarantees.

Abstract

We study non-convex distributed optimization problems where a set of agents collaboratively solve a separable optimization problem that is distributed over a time-varying network. The existing methods to solve these problems rely on (at most) one time-scale algorithms, where each agent performs a diminishing or constant step-size gradient descent at the average estimate of the agents in the network. However, if possible at all, exchanging exact information, that is required to evaluate these average estimates, potentially introduces a massive communication overhead. Therefore, a reasonable practical assumption to be made is that agents only receive a rough approximation of the neighboring agents' information. To address this, we introduce and study a \textit{two time-scale} decentralized algorithm with a broad class of \textit{lossy} information sharing methods (that includes noisy, quantized, and/or compressed information sharing) over \textit{time-varying} networks. In our method, one time-scale suppresses the (imperfect) incoming information from the neighboring agents, and one time-scale operates on local cost functions' gradients. We show that with a proper choices for the step-sizes' parameters, the algorithm achieves a convergence rate of $\mathcal{O}({T}^{-1/3 + \epsilon})$ for non-convex distributed optimization problems over time-varying networks, for any $\epsilon>0$. Our simulation results support the theoretical results of the paper.