Stochastic Coded Federated Learning: Theoretical Analysis and Incentive Mechanism Design

TL;DR

This paper introduces stochastic coded federated learning (SCFL), a framework that uses coded computing and privacy-preserving noise addition to improve training efficiency and privacy, along with an incentive mechanism to balance tradeoffs.

Contribution

The paper proposes a novel SCFL framework combining coded computing with privacy guarantees and designs an incentive mechanism to optimize privacy-performance tradeoff.

Findings

SCFL achieves better model accuracy within fixed training time.

The privacy-performance tradeoff is effectively managed by adjusting noise levels.

The incentive mechanism outperforms traditional Stackelberg approaches.

Abstract

Federated learning (FL) has achieved great success as a privacy-preserving distributed training paradigm, where many edge devices collaboratively train a machine learning model by sharing the model updates instead of the raw data with a server. However, the heterogeneous computational and communication resources of edge devices give rise to stragglers that significantly decelerate the training process. To mitigate this issue, we propose a novel FL framework named stochastic coded federated learning (SCFL) that leverages coded computing techniques. In SCFL, before the training process starts, each edge device uploads a privacy-preserving coded dataset to the server, which is generated by adding Gaussian noise to the projected local dataset. During training, the server computes gradients on the global coded dataset to compensate for the missing model updates of the straggling devices. We design a gradient aggregation scheme to ensure that the aggregated model update is an unbiased estimate of the desired global update. Moreover, this aggregation scheme enables periodical model averaging to improve the training efficiency. We characterize the tradeoff between the convergence performance and privacy guarantee of SCFL. In particular, a more noisy coded dataset provides stronger privacy protection for edge devices but results in learning performance degradation. We further develop a contract-based incentive mechanism to coordinate such a conflict. The simulation results show that SCFL learns a better model within the given time and achieves a better privacy-performance tradeoff than the baseline methods. In addition, the proposed incentive mechanism grants better training performance than the conventional Stackelberg game approach.