BLE-based Indoor Positioning System

Feb 15, 2021 · 1 min read

Master thesis project at KU Leuven in collaboration with NXP Semiconductors.

Compared different Machine Learning algorithms for efficient Bluetooth Low Energy (BLE) based indoor positioning. The system uses Raspberry Pi as signal scanners and NXH3670 for signal advertising.

Key achievements:

  • Implemented both supervised and semi-supervised regression approaches
  • Achieved accurate location prediction using RSSI values
  • Semi-supervised approach showed lower RMSE
Research Machine Learning IoT
Aditya Paliwal
Authors
Aditya Paliwal (he/him)
Data Engineer
Data Engineer with 4+ years of experience in implementing and deploying end-to-end data pipelines in production environments. Passionate about combining data engineering with cutting-edge machine learning and AI technologies to create intelligent, data-driven products.