Scholarlink

In harsh indoor environments, ultra-wideband (UWB) localization system is susceptible to non-line-of-sight (NLOS) effect, which seriously degrades positioning accuracy. This chapter focuses on the application of machine learning (ML) in NLOS identification and mitigation. The UWB channel model is introduced, propagation characteristics of UWB signal are carefully analyzed, and NLOS errors are analyzed in detail by taking through-wall as an example. NLOS identification methods are categorized, and extraction of channel statistical features is analyzed. As an implementation of ML, the application of deep learning (DL) in NLOS identification is also discussed in detail. Furthermore, UWB localization algorithms that take into account NLOS mitigation are studied, and ML-based NLOS error compensation strategies are discussed.

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-Based Non-line-of-sight Identification and Mitigation for Ultra-wideband Localization

  • 动态工具书词条

  • 发表日期：

    2025年

查看全文
  • Feiyang Zhu,
  • Kegen Yu,
  • Jin Wang

摘要

In harsh indoor environments, ultra-wideband (UWB) localization system is susceptible to non-line-of-sight (NLOS) effect, which seriously degrades positioning accuracy. This chapter focuses on the application of machine learning (ML) in NLOS identification and mitigation. The UWB channel model is introduced, propagation characteristics of UWB signal are carefully analyzed, and NLOS errors are analyzed in detail by taking through-wall as an example. NLOS identification methods are categorized, and extraction of channel statistical features is analyzed. As an implementation of ML, the application of deep learning (DL) in NLOS identification is also discussed in detail. Furthermore, UWB localization algorithms that take into account NLOS mitigation are studied, and ML-based NLOS error compensation strategies are discussed.