Scholarlink

<p>In this study, an advanced self-localization system has been developed specifically for NB-IoT networks in indoor environments, employing a low-complexity machine learning approach. To accurately estimate the positions of unknown nodes, a Random Forest model is utilized based on received signal strength indicator (RSSI) measurements from anchor nodes. The system leverages the unique characteristics of NB-IoT signals, such as signal repetitions, to improve localization accuracy while maintaining low computational complexity. A path loss model aligned with the fifth-generation (5&#xa0;G) millimeter-wave (mmWave) standard is incorporated to ensure robustness and precision in localization. Extensive simulations have validated the method’s accuracy, with performance evaluated across diverse system and channel conditions. Results indicate that the proposed model consistently surpasses conventional techniques, achieving an average positioning error of less than 0.4&#xa0;ms. This innovative approach offers considerable potential for enhancing localization capabilities in NB-IoT networks, providing a cost-effective and infrastructure-independent solution for indoor positioning applications.</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

Machine learning-enhanced self-localization for NB-IoT networks in indoor environment

  • 期刊论文

  • 发表日期：

    2025年12月24日

查看全文
  • Anas Alashqar,
  • Ala Khalifeh,
  • Raed Mesleh

摘要

In this study, an advanced self-localization system has been developed specifically for NB-IoT networks in indoor environments, employing a low-complexity machine learning approach. To accurately estimate the positions of unknown nodes, a Random Forest model is utilized based on received signal strength indicator (RSSI) measurements from anchor nodes. The system leverages the unique characteristics of NB-IoT signals, such as signal repetitions, to improve localization accuracy while maintaining low computational complexity. A path loss model aligned with the fifth-generation (5 G) millimeter-wave (mmWave) standard is incorporated to ensure robustness and precision in localization. Extensive simulations have validated the method’s accuracy, with performance evaluated across diverse system and channel conditions. Results indicate that the proposed model consistently surpasses conventional techniques, achieving an average positioning error of less than 0.4 ms. This innovative approach offers considerable potential for enhancing localization capabilities in NB-IoT networks, providing a cost-effective and infrastructure-independent solution for indoor positioning applications.