Multi-Source Data Fusion Method for Indoor Localization System

Jishi Cui,Bin Li,Lyuxiao Yang,Nan Wu
DOI: https://doi.org/10.1109/iccc49849.2020.9238826
2020-01-01
Abstract:In this paper, a multi-source data fusion method for indoor localization system is designed to realize the high-accuracy locations. The indoor localization system consists of WiFi nodes, ultra-wideband (UWB) nodes and inertial measurement unit (IMU), where the IMU is integrated in a android smartphone. For our indoor localization system, there include two stages: offline stage and online stage. In the offline stage, we use crowdsourcing method to train a fingerprint database, which can be constructed at a low labor cost. In the online stage, particle filter to estimate the locations based on WiFi received signal strengths, UWB rangings, and IMU data. Experimental results show that our indoor localization system based on the multi-source data fusion method expands the coverage and improves the localization accuracy.
What problem does this paper attempt to address?