Scholarlink

<p>Cell-free massive MIMO (CF-mMIMO), as a user-centric architecture, significantly enhances the edge user experience rate. Maximizing the downlink sum rate in CF-mMIMO systems critically relies on effective multi-user joint scheduling, pilot allocation and precoding. However, the large number of remote antenna units (RAUs) and users involved can lead to prohibitive computational complexity and excessive data exchange. To address this challenge, this paper proposes a downlink multi-user joint scheduling solution with dynamic user clustering. First, we exploit the spatial sparsity of the user channel matrix to build scalable dynamic user clusters. Scheduling is then performed using clusters as basic units, with real-time intra-cluster and progressive inter-cluster processing, reducing computational complexity. Second, based on the clustering results, we design a low-complexity, high-performance multi-user joint time-frequency resource allocation scheme. Third, we develop an adaptive pilot allocation method for the scheduling framework that rivals capacity-maximizing pilot reuse strategies. Finally, we optimize zero-forcing to construct real-time intra-cluster and statistical inter-cluster interference cancellation weights. The resulting concatenated precoding weights offer superior interference suppression. Simulation results validate that the proposed joint scheduling scheme achieves a higher sum rate while maintaining comparable user fairness. Moreover, the optimized precoding algorithm delivers a significant rate gain exceeding 12%.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Dynamic User Cluster-Based Downlink Multi-user Joint Scheduling and Precoding for Cell-Free Massive MIMO Systems

  • 期刊论文

  • 发表日期：

    2025年4月1日

查看全文
  • Dekun Zhang,
  • Baoming Bai

摘要

Cell-free massive MIMO (CF-mMIMO), as a user-centric architecture, significantly enhances the edge user experience rate. Maximizing the downlink sum rate in CF-mMIMO systems critically relies on effective multi-user joint scheduling, pilot allocation and precoding. However, the large number of remote antenna units (RAUs) and users involved can lead to prohibitive computational complexity and excessive data exchange. To address this challenge, this paper proposes a downlink multi-user joint scheduling solution with dynamic user clustering. First, we exploit the spatial sparsity of the user channel matrix to build scalable dynamic user clusters. Scheduling is then performed using clusters as basic units, with real-time intra-cluster and progressive inter-cluster processing, reducing computational complexity. Second, based on the clustering results, we design a low-complexity, high-performance multi-user joint time-frequency resource allocation scheme. Third, we develop an adaptive pilot allocation method for the scheduling framework that rivals capacity-maximizing pilot reuse strategies. Finally, we optimize zero-forcing to construct real-time intra-cluster and statistical inter-cluster interference cancellation weights. The resulting concatenated precoding weights offer superior interference suppression. Simulation results validate that the proposed joint scheduling scheme achieves a higher sum rate while maintaining comparable user fairness. Moreover, the optimized precoding algorithm delivers a significant rate gain exceeding 12%.