Bandwidth Slicing to Boost Federated Learning in Edge Computing

TL;DR

This paper proposes bandwidth slicing as a technique to enhance federated learning efficiency in edge computing by reducing communication delays, leading to faster training without sacrificing accuracy.

Contribution

It introduces bandwidth slicing specifically tailored for federated learning in edge environments, improving training speed and efficiency.

Findings

Bandwidth slicing significantly reduces communication delay.

Training efficiency is notably improved with bandwidth slicing.

Learning accuracy remains high despite reduced communication delays.

Abstract

Bandwidth slicing is introduced to support federated learning in edge computing to assure low communication delay for training traffic. Results reveal that bandwidth slicing significantly improves training efficiency while achieving good learning accuracy.