Knowledge Distillation for mmWave Beam Prediction Using Sub-6 GHz Channels

TL;DR

This paper introduces a lightweight knowledge distillation framework that enables efficient mmWave beam prediction using sub-6 GHz channels, significantly reducing model complexity while maintaining high accuracy.

Contribution

It develops two compact student deep learning models based on distillation strategies that match large teacher models' performance with 99% fewer parameters.

Findings

Student models achieve teacher-level beam prediction accuracy.

Spectral efficiency comparable to large models.

Parameter and computational complexity reduced by 99%.

Abstract

Beamforming in millimeter-wave (mmWave) high-mobility environments typically incurs substantial training overhead. While prior studies suggest that sub-6 GHz channels can be exploited to predict optimal mmWave beams, existing methods depend on large deep learning (DL) models with prohibitive computational and memory requirements. In this paper, we propose a computationally efficient framework for sub-6 GHz channel-mmWave beam mapping based on the knowledge distillation (KD) technique. We develop two compact student DL architectures based on individual and relational distillation strategies, which retain only a few hidden layers yet closely mimic the performance of large teacher DL models. Extensive simulations demonstrate that the proposed student models achieve the teacher's beam prediction accuracy and spectral efficiency while reducing trainable parameters and computational complexity by 99%.