Scholarlink

In 5G networks, reliable data transmission for massive machine-type communication (mMTC) devices hinges on accurate channel prediction. This paper introduces a novel approach based on channel prediction for resource allocation in 5G mMTC using graph neural network (AMC-mMTC-GCIGNN). The method leverages granger causality-inspired graph neural networks to enhance channel prediction accuracy by analyzing feature relationships within channel data. Comparative analysis against existing techniques using performance metrics like bit error rate, mean squared error, and signal-to-noise ratio highlights the superiority of the proposed approach. The results underscore its potential to significantly enhance communication performance within 5G mMTC systems, thereby addressing a crucial aspect of next-generation wireless networks.

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

Channel Prediction for Resource Allocation in 5G Massive Machine-Type Communications Using Graph Neural Network

  • 会议论文

  • 发表日期：

    2025年

查看全文
  • Nishu Gupta,
  • Jukka Mäkelä,
  • Mikko Uitto,
  • Arun Prakash

摘要

In 5G networks, reliable data transmission for massive machine-type communication (mMTC) devices hinges on accurate channel prediction. This paper introduces a novel approach based on channel prediction for resource allocation in 5G mMTC using graph neural network (AMC-mMTC-GCIGNN). The method leverages granger causality-inspired graph neural networks to enhance channel prediction accuracy by analyzing feature relationships within channel data. Comparative analysis against existing techniques using performance metrics like bit error rate, mean squared error, and signal-to-noise ratio highlights the superiority of the proposed approach. The results underscore its potential to significantly enhance communication performance within 5G mMTC systems, thereby addressing a crucial aspect of next-generation wireless networks.