Scholarlink

A smooth and optimal utilization of mmWave and sub-6 GHz IoT networks with more mobile users is necessary to satisfy the demanding requirements of Beyond 5G applications. This manuscript specifically suggests a Multiscale Deep Convolutional Neural Network (MSDCNN) for Improved Beam Prediction in Sub-6GHz Channels that forecasts the mmWave obstruction condition in the future and the mobile user’s optimal beamforming vectors based only on knowledge of sub-6 GHz channels or out-of-band data, to lessen the significant burdens associated with mmWave channel state information feedback. In this manuscript, utilizing a minuscule amount of uplink sub-6GHz channel data and mmWave pilots, a unique downstream beamforming technique for mmWave communications is proposed. The suggested model offers 21.04%, 15.47%, and 9.63% reduced signal-to-noise ratio (SNR) compared to current methods.

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

Multiscale Deep Convolutional Neural Network for Enhanced Beam Prediction for mmWave Communications

  • 会议论文

  • 发表日期：

    2026年

查看全文
  • M. M. Kamruzzaman

摘要

A smooth and optimal utilization of mmWave and sub-6 GHz IoT networks with more mobile users is necessary to satisfy the demanding requirements of Beyond 5G applications. This manuscript specifically suggests a Multiscale Deep Convolutional Neural Network (MSDCNN) for Improved Beam Prediction in Sub-6GHz Channels that forecasts the mmWave obstruction condition in the future and the mobile user’s optimal beamforming vectors based only on knowledge of sub-6 GHz channels or out-of-band data, to lessen the significant burdens associated with mmWave channel state information feedback. In this manuscript, utilizing a minuscule amount of uplink sub-6GHz channel data and mmWave pilots, a unique downstream beamforming technique for mmWave communications is proposed. The suggested model offers 21.04%, 15.47%, and 9.63% reduced signal-to-noise ratio (SNR) compared to current methods.