mmWave Beam Selection in Analog Beamforming Using Personalized Federated Learning

TL;DR

This paper introduces a personalized federated learning approach with FedLion for mmWave beam selection, effectively handling non-IID data to improve beam configuration accuracy and reduce signaling overhead.

Contribution

It proposes a novel personalized federated learning method and FedLion optimizer for mmWave beam selection in heterogeneous scenarios.

Findings

Up to 33.6% higher accuracy than a single FL model

6% higher accuracy than a local model

Reduces signaling overhead

Abstract

Using analog beamforming in mmWave frequency bands we can focus the energy towards a receiver to achieve high throughput. However, this requires the network to quickly find the best downlink beam configuration in the face of non-IID data. We propose a personalized Federated Learning (FL) method to address this challenge, where we learn a mapping between uplink Sub-6GHz channel estimates and the best downlink beam in heterogeneous scenarios with non-IID characteristics. We also devise FedLion, a FL implementation of the Lion optimization algorithm. Our approach reduces the signaling overhead and provides superior performance, up to 33.6% higher accuracy than a single FL model and 6% higher than a local model.