Scholarlink

<p>Massive multiple inputs and multiple outputs (M-MIMO) is one of the vital technologies for systems beyond fifth-generation (B5G). The signal detection 256 × 256 M-MIMO is complex and expensive due to the utilization of several antennas' arrays. In this letter, we give a Graph Neural Network (GNN) detection to investigate the bit error rate (BER), power spectral density (PSD), and complexity performance of the M-MIMO system under perfect channel state information (CSI) and imperfect CSI. The simulation results reveal that the proposed GNN achieved a 63% power-saving performance and also reduced the slide lobes to − 710&#xa0;dB and − 590&#xa0;dB, outperforming the conventional schemes.</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

Graph Neural Network Signal Detection for 256 × 256 MIMO Under Imperfect Channel State Information

  • 期刊论文

  • 发表日期：

    2026年4月9日

查看全文
  • Arun Kumar,
  • Nishant Gaur,
  • Aziz Nanthaamornphong

摘要

Massive multiple inputs and multiple outputs (M-MIMO) is one of the vital technologies for systems beyond fifth-generation (B5G). The signal detection 256 × 256 M-MIMO is complex and expensive due to the utilization of several antennas' arrays. In this letter, we give a Graph Neural Network (GNN) detection to investigate the bit error rate (BER), power spectral density (PSD), and complexity performance of the M-MIMO system under perfect channel state information (CSI) and imperfect CSI. The simulation results reveal that the proposed GNN achieved a 63% power-saving performance and also reduced the slide lobes to − 710 dB and − 590 dB, outperforming the conventional schemes.