Decentralized Federated Averaging

TL;DR

This paper introduces a decentralized federated averaging algorithm with momentum and quantization, reducing communication costs and enhancing privacy by eliminating the central server, with proven convergence and empirical validation.

Contribution

It proposes a novel decentralized FedAvg variant with momentum and quantization, improving communication efficiency and privacy in federated learning.

Findings

Convergence of the proposed decentralized algorithm is theoretically proven.

Quantized version maintains convergence with reduced communication.

Numerical experiments demonstrate the effectiveness of the method.

Abstract

Federated averaging (FedAvg) is a communication efficient algorithm for the distributed training with an enormous number of clients. In FedAvg, clients keep their data locally for privacy protection; a central parameter server is used to communicate between clients. This central server distributes the parameters to each client and collects the updated parameters from clients. FedAvg is mostly studied in centralized fashions, which requires massive communication between server and clients in each communication. Moreover, attacking the central server can break the whole system's privacy. In this paper, we study the decentralized FedAvg with momentum (DFedAvgM), which is implemented on clients that are connected by an undirected graph. In DFedAvgM, all clients perform stochastic gradient descent with momentum and communicate with their neighbors only. To further reduce the communication cost, we also consider the quantized DFedAvgM. We prove convergence of the (quantized) DFedAvgM under trivial assumptions; the convergence rate can be improved when the loss function satisfies the P{\L} property. Finally, we numerically verify the efficacy of DFedAvgM.