Sparsification of Precoding Codebooks for PAPR Reduction via Grassmannian Representations
Joe Asano, Yuto Hama, Hiroki Iimori · 2026
Read Sparsification of Precoding Codebooks for PAPR Reduction via Grassmannian Representations by Joe Asano, Yuto Hama, Hiroki Iimori (2026) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2604-19273.txt.
Introduction
Asano and colleagues propose thinning out precoding codebooks, the sets of beamforming patterns used in wireless transmission, to lower peak-to-average power ratio without losing data rate. Because precoder matrices sit on the Grassmann manifold, the codebook can be thinned without changing the feedback mechanism. The goal is lower peaks with no real loss in speed or quality. They use exact thinning by unitary transformation and approximate thinning by sparse principal component analysis. The thinned codebooks cut peak power by more than 1 dB in uplink tests at negligible cost.
Catalog record
| CREATOR | Joe Asano, Yuto Hama, Hiroki Iimori |
|---|---|
| YEAR | 2026 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 28146 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2604.19273
Browse the full library · About the Tower of Babel
Full text
The opening of the work, from the library's hosted text:
Sparsification of Precoding Codebooks for PAPR Reduction via Grassmannian Representations Thanks: J. Asano and N. Ishikawa are with the Faculty of Engineering, Yokohama National University, 240-8501 Kanagawa, Japan (e-mail: [email protected]). Y. Hama, H. Iimori, and S. Malomsoky are with Ericsson Research, Ericsson Japan K. K., Yokohama, 220-0012 Kanagawa, Japan (e-mail: [yuto.hama, hiroki.iimori, szabolcs.malomsoky]@ericsson.com). Joe Asano, Yuto Hama, Hiroki Iimori 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 (SPCA), and integrate them into a unified design algorithm. The proposed sparsified codebooks incur negligible performance loss while reducing PAPR by more than 1 dB 1~\text{dB} in uplink scenarios. I Introduction IN the evolution toward 6G systems, uplink traffic is expected to increase significantly due to the growing number of connected devices [ 5 ] . In uplink systems, power amplifier (PA) efficiency is inherently constrained at user equipment, and the high peak-to-average power ratio (PAPR) remains a major challenge in orthogonal frequency-division multiplexing (OFDM) signals, which are commonly used in modern wireless communication standards. Extensive studies have characterized the statistical properties of PAPR and developed reduction techniques for OFDM and discrete Fourier transform spread OFDM (DFT-s-OFDM) [ 10 , 7 ] , which are adopted in current 5G uplink systems. To further enhance both uplink capacity and coverage in beyond-5G and 6G systems, extending DFT-s-OFDM to support multi-stream transmission has attracted increasing interest, since the current 5G standard restricts it to single-stream transmission. Spatial precoding plays an essential role in multiple-input multiple-output (MIMO) systems. In 5G uplink, codebook-based precoding is employed, where the codebooks composed of multiple precoders are standardized in [ 1 ] for different numbers of antenna ports and data streams. The precoding matrices can be interpreted as points on a Grassmann manifold [ 8 , 9 ] , and under uncorrelated Rayleigh fading, codebook design is equivalent to subspace packing on the Grassmann manifold. However, existing codebooks adopt dense precoder structures designed to maximize array gain and their design does not take into account the use of multi-stream DFT-s-OFDM. When applied to multi-stream transmission, such dense precoders break the single-carrier property of DFT-s-OFDM, resulting in increased PAPR. Recent work [ 2 ] addressed this issue by designing codebooks consisting exclusively of sparse precoders and demonstrated their effectiveness in reducing PAPR for multi-stream DFT-s-OFDM. In this letter, we propose a sparsification method for precoding codebooks that reduces PAPR while maintaining the achievable rate. Unlike [ 2 ] , which designs new codebooks, the proposed method sparsifies existing codebooks standardized in [ 1 ] while preserving the original codebook indices. This approach enables effective PAPR reduction without affecting existing standard specifications, since dense precoders can be replaced with sparse precoders through a user-equipment-only implementation. By exploiting the Grassmannian representation of precoder matrices, sparse precoders are constructed either via unitary transformations of the original precoders or via sparse principal component analysis (SPCA) [ 3 , 11 ] . The main contributions are summarized as follows: 1. We propose a codebook-based precoding framework that uses sparse precoders instead of dense precoders for PAPR reduction in DFT-s-OFDM without performance degradation and without modifying the existing feedback mechanism. 2. We develop two closed-form sparsification methods on the Grassmann manifold, based on unitary transformations and SPCA. II System Model In this section, we describe the limited-feedback MIMO system [ 8 , 9 ] with codebook-based precoding. We consider a single-user uplink limited-feedback MIMO system with N_{t} transmit and N_{r} receive antennas, where N_{s}\leq\min(N_{t},N_{r}) denotes the number of data streams. Let \mathbf{H}_{k}\in\mathbb{C}^{N_{r}\times N_{t}} denote the channel matrix for the k -th subcarrier, k=1,\ldots,K . Let \mathbf{s}_{k}\in\mathbb{C}^{N_{s}\times 1} denote the transmitted symbol vector on the k -th subcarrier, drawn from phase shift keying (PSK) or quadrature amplitude modulation (QAM) constellations and satisfying \mathbb{E}[\mathbf{s}_{k}\mathbf{s}_{k}^{\mathrm{H}}]=\mathbf{I}_{N_{s}} . The received signal \mathbf{y}_{k}\in\mathbb{C}^{N_{r}\times 1} is given by \displaystyle\mathbf{y}_{k}=\mathbf{H}_{k}\mathbf{W}_{i}\mathbf{s}_{k}+\mathbf{v}_{k}, (1) where \mathbf{W}_{i}\in\mathbb{C}^{N_{t}\times N_{s}} is selected from a finite codebook \mathcal{W}={\mathbf{W}_{1},\ldots,\mathbf{W}_{|\mathcal{W}|}} , \mathbf{H}_{k} is the channel matrix, and \mathbf{v}_{k}\sim\mathcal{CN}(\mathbf{0},\sigma_{v}^{2}\mathbf{I}_{N_{r}}) is additive white Gaussian noise. We consider wideband precoding as in current 5G uplink systems, where a common precoder is applied across all subcarriers. Each precoder satisfies \mathbf{W}_{i}^{\mathrm{H}}\mathbf{W}_{i}=\mathbf{I}_{N_{s}} . With perfect channel state information at the receiver, the precoder is selected to maximize the achievable rate \displaystyle R(\mathbf{W}_{i})=\sum_{k=1}^{K}~\log_{2}\det\left(\mathbf{I}_{N_{s}}+\frac{\rho}{N_{s}}\mathbf{W}_{i}^{\mathrm{H}}\mathbf{H}_{k}^{\mathrm{H}}\mathbf{H}_{k}\mathbf{W}_{i}\right), (2) where \rho=1/\sigma_{v}^{2} denotes the average signal-to-noise ratio (SNR). As seen from ( 2 ), R(\mathbf{W}_{i}) is invariant under right multiplication of \mathbf{W}_{i} by any unitary matrix. Together with the orthonormality constraint, this implies that \mathbf{W}_{i} can be regarded as a point on the Grassmann manifold \mathcal{G}(N_{t},N_{s}) . The receiver feeds back the index \hat{i} corresponding to the optimal precoder as \displaystyle\hat{i}=\underset{1\leq i\leq|\mathcal{W}|}{\text{argmax}}~R(\mathbf{W}_{i}), (3) using \lceil\log_{2}|\mathcal{W}|\rceil bits. III Proposed Sparsification Methods In this section, we describe a new problem formulation for reducing PAPR without modifying the existing feedback mechanism and the corresponding codebook design procedure. III-A Problem Formulation To reduce PAPR in multi-stream DFT-s-OFDM, we propose a pre-mapped sparse codebook design that operates within the existing limited-feedback framework. An overview of the system is summarized as follows: • The UE independently selects a precoder from the pre-mapped sparse codebook using the same feedback index as the original dense codebook. • The selected precoder reduces PAPR with little performance loss in multi-stream DFT-s-OFDM due to its sparse structure. In this system, the pre-mapped sparse codebook should preserve the performance and spatial characteristics (e.g., main lobe directions) of the original dense precoder at each index, since the UE independently selects the precoder without modifying the feedback mechanism. From the viewpoint of codebook design, each precoder represents a point on the Grassmann manifold \mathcal{G}(N_{t},N_{s}) . Hence, the performance preservation problem can be interpreted as approximating each original precoder by a sparse precoder on the Grassmann manifold. On the Grassmann manifold, the distance between two points is defined as the chordal distance. Specifically, the chordal distance between two points \mathbf{W} and \mathbf{P} , denoted by d_{\mathrm{c}}(\mathbf{W},\mathbf{P}) , is defined as \displaystyle d_{\mathrm{c}}(\mathbf{W},\mathbf{P})=\sqrt{N_{s}-\|\mathbf{W}^{\mathrm{H}}\mathbf{P}\|^{2}_{\mathrm{F}}}. (4) Let the original dense codebook be denoted as \mathcal{W}=\{\mathbf{W}_{1},\cdots,\mathbf{W}_{|\mathcal{W}|}\} and the pre-mapped sparse codebook as \mathcal{P}=\{\mathbf{P}_{1},\cdots,\mathbf{P}_{|\mathcal{P}|}\} . Based on the results demonstrated in [ 2 ] , we set the number of nonzero elements of each precoder matrix \mathbf{P}_{i} to N_{t} , i.e., \|\mathbf{P}_{i}\|_{0}=N_{t} , taking into account the trade-off between sparsity and performance. Since the optimization can be carried out independently for each index i\in\{1,\cdots,|\mathcal{W}|\} , the pre-mapped sparse codebook design reduces to the following per-index problem: \displaystyle\begin{split}\underset{\mathbf{P}}{\operatorname{minimize}}\;&d_{\mathrm{c}}(\mathbf{W},\mathbf{P})\\ \text{s.t.}\;&\mathbf{P}\in\mathcal{G}(N_{t},N_{s}),\;\|\mathbf{P}\|_{0}=N_{t}.\end{split} (5) Thus, the index i is omitted hereafter without loss of generality, since each precoder is optimized independently. III-B Exact Sparsification via Unitary Transformation Regarding the problem defined in ( 5 ), when the optimal sparse precoder \mathbf{P} exists, that is, the chordal distance between \mathbf{W} and \mathbf{P} is 0 , they span the same subspace, and there exists a unitary matrix \mathbf{U}\in\mathbb{C}^{N_{s}\times N_{s}} such that \displaystyle\mathbf{P}=\mathbf{W}\mathbf{U}. (6) In this work, we demonstrate that the derivation of such a unitary matrix \mathbf{U} and corresponding sparse precoder