Neural Collapse based Deep Supervised Federated Learning for Signal Detection in OFDM Systems

TL;DR

This paper introduces a neural collapse inspired deep supervised federated learning algorithm to improve signal detection in OFDM systems within AI-empowered wireless networks, addressing data heterogeneity issues.

Contribution

The paper proposes a novel NCDSFL algorithm that leverages neural collapse principles to enhance federated learning for wireless signal detection.

Findings

Improved accuracy in signal detection under data heterogeneity

Enhanced robustness of federated learning in wireless environments

Demonstrated effectiveness through simulations

Abstract

Future wireless networks are expected to be AI-empowered, making their performance highly dependent on the quality of training datasets. However, physical-layer entities often observe only partial wireless environments characterized by different power delay profiles. Federated learning is capable of addressing this limited observability, but often struggles with data heterogeneity. To tackle this challenge, we propose a neural collapse (NC) inspired deep supervised federated learning (NCDSFL) algorithm.