On In-network learning. A Comparative Study with Federated and Split Learning

TL;DR

This paper introduces 'in-network learning,' a new architecture for distributed inference in wireless networks, demonstrating it outperforms federated and split learning in accuracy and bandwidth efficiency.

Contribution

The paper proposes a novel in-network learning architecture, including a specific loss function and neural network optimization, and compares its performance with existing federated and split learning methods.

Findings

In-network learning achieves higher accuracy than federated and split learning.

It offers significant bandwidth savings compared to existing methods.

Experimental results validate the effectiveness of the proposed architecture.

Abstract

In this paper, we consider a problem in which distributively extracted features are used for performing inference in wireless networks. We elaborate on our proposed architecture, which we herein refer to as "in-network learning", provide a suitable loss function and discuss its optimization using neural networks. We compare its performance with both Federated- and Split learning; and show that this architecture offers both better accuracy and bandwidth savings.