Sparsification of Precoding Codebooks for PAPR Reduction via Grassmannian Representations

TL;DR

This paper introduces a novel sparsification technique for precoding codebooks on the Grassmann manifold, significantly reducing PAPR with minimal performance loss.

Contribution

It presents two new sparsification methods—exact and approximate—that preserve feedback mechanisms and improve PAPR reduction in wireless systems.

Findings

PAPR reduced by more than 1 dB in uplink scenarios.

Negligible performance loss with the proposed sparsification methods.

Unified algorithm effectively integrates both sparsification approaches.

Abstract

In this letter, we propose a sparsification method for precoding codebooks that reduces the peak-to-average power ratio (PAPR) while preserving the achievable rate. By exploiting the fact that precoder matrices lie on the Grassmann manifold, we formulate a codebook design problem that enables sparsification without modifying the existing feedback mechanism. We develop two sparsification approaches, namely exact sparsification via unitary transformation and approximate sparsification via sparse principal component analysis, and integrate them into a unified design algorithm. The proposed sparsified codebooks incur negligible performance loss while reducing PAPR by more than 1 dB in uplink scenarios.