Serving Federated Learning and Non-Federated Learning Users: A Massive MIMO Approach

TL;DR

This paper proposes a massive MIMO-based communication scheme to efficiently serve both federated learning and non-federated learning users, optimizing resource allocation for improved data rates and quality of service.

Contribution

It introduces a novel resource allocation scheme for massive MIMO systems that jointly serves FL and non-FL users, with an optimization algorithm for power control.

Findings

Significant performance improvement over baseline schemes

Effective optimization of transmit power for fairness and QoS

Successful application of successive convex approximation algorithm

Abstract

Federated learning (FL) with its data privacy protection and communication efficiency has been considered as a promising learning framework for beyond-5G/6G systems. We consider a scenario where a group of downlink non-FL users are jointly served with a group of FL users using massive multiple-input multiple-output technology. The main challenge is how to utilise the resource to optimally serve both FL and non-FL users. We propose a communication scheme that serves the downlink of the non-FL users (UEs) and the uplink of FL UEs in each half of the frequency band. We formulate an optimization problem for optimizing transmit power to maximize the minimum effective data rates for non-FL users, while guaranteeing a quality-of-service time of each FL communication round for FL users. Then, a successive convex approximation-based algorithm is proposed to solve the formulated problem. Numerical results confirm that our proposed scheme significantly outperforms the baseline scheme.