Scholarlink

In access-controlled environments, indoor positioning faces complications due to feature interference, multi-path propagation, and environmental reflections, making traditional techniques (RSSI, AoA, ToA) less accurate. This work proposes a Cabin-Finding Application that employs Bluetooth channel sounding and advanced distance measurement techniques (MCPD, IFFT, Phase Slope, and RSSI) to achieve improved accuracy. The proposed system integrates a hybrid machine learning model combining XGBoost, LightGBM, CatBoost, and a neural network meta-learner to refine distance estimations and reduce discrepancies. The system collects real-time channel data from nRF52840 Bluetooth 5.3 devices and offers precise indoor navigation via an interactive map-based UI. Experimental comparisons demonstrate that the proposed system significantly outperforms conventional BLE-based indoor positioning methods, highlighting the potential of Bluetooth channel sounding combined with machine learning for accurate and scalable indoor localization.

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

Indoor Navigation System with Faculty Cabin Identification Using Bluetooth Channel Sounding and Distance Measurement APIs

  • 会议论文

  • 发表日期：

    2026年

查看全文
  • Harsha Kusu,
  • Geeth Vishnu Gandodi,
  • Jaswanth Uttaravelli,
  • Gowtham Veluguri,
  • J. Govindarajan

摘要

In access-controlled environments, indoor positioning faces complications due to feature interference, multi-path propagation, and environmental reflections, making traditional techniques (RSSI, AoA, ToA) less accurate. This work proposes a Cabin-Finding Application that employs Bluetooth channel sounding and advanced distance measurement techniques (MCPD, IFFT, Phase Slope, and RSSI) to achieve improved accuracy. The proposed system integrates a hybrid machine learning model combining XGBoost, LightGBM, CatBoost, and a neural network meta-learner to refine distance estimations and reduce discrepancies. The system collects real-time channel data from nRF52840 Bluetooth 5.3 devices and offers precise indoor navigation via an interactive map-based UI. Experimental comparisons demonstrate that the proposed system significantly outperforms conventional BLE-based indoor positioning methods, highlighting the potential of Bluetooth channel sounding combined with machine learning for accurate and scalable indoor localization.