Efficient Subcarrier Grouping Methods for Hybrid Precoding in Massive MIMO-OFDM Systems

Abstract

This paper presents k-means based novel static and dynamic subcarrier grouping algorithms for hybrid precoding in broadband massive MIMO systems, thereby addressing critical limitations of conventional clustering methods. The primary innovation of this work is the implementation of eigenvector-based correlation analysis of the channel to optimize dynamic subcarrier grouping. This approach has been demonstrated to enhance both the bit error rate performance and the achievable rate under frequency-selective channels. In order to compare the performance of the proposed methods with the theoretical lower bounds, approximate closed-form bounds are obtained for both the BER expression and the achievable rate expression. © 2025 Elsevier B.V., All rights reserved.