Scholarlink

<p>The ubiquitous use of drones in a broad spectrum of fields, particularly in farming and agriculture, surveillance, and delivery services not only presents significant opportunities but also raises safety and security concerns. Therefore, to ensure a safe and secure operation, effective management and critical monitoring of drone activity are essential. However, the conventional drone detection approaches that include radar and visual surveillance are constrained to limitations related to accuracy, range and environmental susceptibility. Hence, to address these shortcomings, this study focuses on 5GDC that fuses 5G Channel State Information (CSI) with advanced machine learning approaches to approximate the drone count in a particular scene. The underlying high-resolution and precision properties of 5G technology enable CSI data to be effectively used for drone detection and identification. A multichannel 1D Convolutional Neural Network (1D-CNN) is employed in this work to analyze the one-dimensional time series data from CSI measurements. The proposed neural network model automatically learns and extracts the features indicative of drone presence and movement, improving detection accuracy and reliability. Finally, the performance of the 5GDC method is then evaluated through numerical results.</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

5GDC-estimating drone count using 5G CSI measurements and multi-channel CNN

  • 期刊论文

  • 出版模式：

    开放获取

  • 发表日期：

    2025年12月13日

查看全文
  • Sriram Mahateja Akella,
  • Priya Thangaraj,
  • Sreenivasa Reddy Yeduri,
  • Linga Reddy Cenkeramaddi

摘要

The ubiquitous use of drones in a broad spectrum of fields, particularly in farming and agriculture, surveillance, and delivery services not only presents significant opportunities but also raises safety and security concerns. Therefore, to ensure a safe and secure operation, effective management and critical monitoring of drone activity are essential. However, the conventional drone detection approaches that include radar and visual surveillance are constrained to limitations related to accuracy, range and environmental susceptibility. Hence, to address these shortcomings, this study focuses on 5GDC that fuses 5G Channel State Information (CSI) with advanced machine learning approaches to approximate the drone count in a particular scene. The underlying high-resolution and precision properties of 5G technology enable CSI data to be effectively used for drone detection and identification. A multichannel 1D Convolutional Neural Network (1D-CNN) is employed in this work to analyze the one-dimensional time series data from CSI measurements. The proposed neural network model automatically learns and extracts the features indicative of drone presence and movement, improving detection accuracy and reliability. Finally, the performance of the 5GDC method is then evaluated through numerical results.