AI Chat AI Image Generator AI Video Text to Speech

Optimal MIMO Combining for Blind Federated Edge Learning with Gradient Sparsification

03/24/2022

∙

by Ema Becirovic, et al.

∙

∙

We provide the optimal receive combining strategy for federated learning in multiple-input multiple-output (MIMO) systems. Our proposed algorithm allows the clients to perform individual gradient sparsification which greatly improves performance in scenarios with heterogeneous (non i.i.d.) training data. The proposed method beats the benchmark by a wide margin.

Ema Becirovic
3 publications
Zheng Chen
71 publications
Erik G. Larsson
102 publications

page 1

page 2

page 3

page 4

research

∙ 06/12/2022

Communication-Efficient Federated Learning over MIMO Multiple Access Channels

Communication efficiency is of importance for wireless federated learnin...

0 Yo-Seb Jeon, et al. ∙

research

∙ 06/03/2023

Over-the-Air Federated Learning In Broadband Communication

Federated learning (FL) is a privacy-preserving distributed machine lear...

0 Wayne Lemieux, et al. ∙

research

∙ 03/18/2020

Gradient Estimation for Federated Learning over Massive MIMO Communication Systems

Federated learning is a communication-efficient and privacy-preserving s...

0 Yo-Seb Jeon, et al. ∙

research

∙ 01/31/2023

Federated Cell-Free MIMO in Non-Terrestrial Networks: Architectures and Performance

While 5G networks are being rolled out, the definition of the potential ...

0 Alessandro Guidotti, et al. ∙

research

∙ 01/29/2022

Random Orthogonalization for Federated Learning in Massive MIMO Systems

We propose a novel uplink communication method, coined random orthogonal...

0 Xizixiang Wei, et al. ∙

research

∙ 12/17/2021

Federated Learning with Heterogeneous Data: A Superquantile Optimization Approach

We present a federated learning framework that is designed to robustly d...

13 Krishna Pillutla, et al. ∙

research

∙ 09/15/2021

Federated Learning of Molecular Properties in a Heterogeneous Setting

Chemistry research has both high material and computational costs to con...

15 Wei Zhu, et al. ∙