Scholarlink

The advancement of sixth-generation (6G) wireless networks has opened new horizons for Integrated Sensing and Communication (ISAC), particularly in the detection of Unmanned Aerial Vehicles (UAVs). This paper presents a novel ISAC-based framework that integrates deep learning with THz communications, beamforming, and massive MIMO to estimate drone altitudes at multiple flight levels accurately. We develop and train a Long Short-Term Memory (LSTM) deep learning network using simulated data, incorporating Doppler frequency, Signal-to-Interference-Plus-Noise Ratio (SINR), and distance as input features. The proposed model achieves high prediction accuracy, with an MSE of 0.0023 and R2 of 0.94. Experimental results reveal a strong correlation between SINR and prediction accuracy, as well as between Doppler frequency and drone altitude. Our findings demonstrate the potential of AI-powered 6G ISAC systems for efficient drone detection in future smart surveillance 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

An Integrated Sensing and Communication Deep Learning Technique in 6G Network for Drone Detection at Multiple Altitudes

  • 会议论文

  • 发表日期：

    2026年

查看全文
  • Shaymaa A. Kashmoola,
  • Ali Othman Al Janaby,
  • Hussain al Rizzo,
  • Manar Kashmula

摘要

The advancement of sixth-generation (6G) wireless networks has opened new horizons for Integrated Sensing and Communication (ISAC), particularly in the detection of Unmanned Aerial Vehicles (UAVs). This paper presents a novel ISAC-based framework that integrates deep learning with THz communications, beamforming, and massive MIMO to estimate drone altitudes at multiple flight levels accurately. We develop and train a Long Short-Term Memory (LSTM) deep learning network using simulated data, incorporating Doppler frequency, Signal-to-Interference-Plus-Noise Ratio (SINR), and distance as input features. The proposed model achieves high prediction accuracy, with an MSE of 0.0023 and R2 of 0.94. Experimental results reveal a strong correlation between SINR and prediction accuracy, as well as between Doppler frequency and drone altitude. Our findings demonstrate the potential of AI-powered 6G ISAC systems for efficient drone detection in future smart surveillance networks.