Scholarlink

Spectrum sensing is essential for future wireless communications, distinguishing 5G, Long-Term Evolution (LTE), and noise signals. It ensures effective coexistence by identifying and allocating available spectrum for diverse technologies. This capability optimizes network performance, mitigates interference, and facilitates seamless integration of IoT devices, enhancing overall communication efficiency and reliability. Recently deep learning (DL) architectures are vital for classifying 5G new radio (5G NR), LTE, and noise signals in wireless communication. They automate feature extraction, improving accuracy and spectrum management for seamless coexistence of diverse technologies. The paper presents two deep neural network architectures, Convolutional Neural Network (CNN) and VGG16, designed to discern between 5G NR and LTE signals. The primary objective is to enhance spectrum awareness by proficiently identifying distinct signal patterns within a cellular system environment. A comprehensive performance analysis of classifiers is conducted, leveraging with different optimizers. Additionally, the research investigates the impact of varying training rates on the classifiers’ efficacy, contributing insights into their comparative superiority.

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

Enhancing Spectrum Awareness in Cellular Networks Through Deep Learning Approaches for Efficient 5G-NR and LTE Signal Classification

  • 会议论文

  • 发表日期：

    2026年

查看全文
  • M. Mani Kumar Reddy,
  • M. Monisha

摘要

Spectrum sensing is essential for future wireless communications, distinguishing 5G, Long-Term Evolution (LTE), and noise signals. It ensures effective coexistence by identifying and allocating available spectrum for diverse technologies. This capability optimizes network performance, mitigates interference, and facilitates seamless integration of IoT devices, enhancing overall communication efficiency and reliability. Recently deep learning (DL) architectures are vital for classifying 5G new radio (5G NR), LTE, and noise signals in wireless communication. They automate feature extraction, improving accuracy and spectrum management for seamless coexistence of diverse technologies. The paper presents two deep neural network architectures, Convolutional Neural Network (CNN) and VGG16, designed to discern between 5G NR and LTE signals. The primary objective is to enhance spectrum awareness by proficiently identifying distinct signal patterns within a cellular system environment. A comprehensive performance analysis of classifiers is conducted, leveraging with different optimizers. Additionally, the research investigates the impact of varying training rates on the classifiers’ efficacy, contributing insights into their comparative superiority.