Enhanced Indoor Localization Based BLE Using Gaussian Process Regression and Improved Weighted kNN

Indoor positioning has attracted commercial developers and researchers in the last few decades. Global system (GPS) cannot well localize indoor environment. For clinical applications such as workflow analysis, Bluetooth Low Energy (BLE) beacons have been employed to map positions of individuals environments. The received signal strength indicator (RSSI) fingerprinting plays a key role access point performance services. This paper deals with issue localization based RSSI fingerprint using only vectors without any prior knowledge pose. We proposed use machine learning (ML) combined modified kNN algorithm enhance real-time propose method detect uncertainty estimated mentioned ML is Gaussian Process Regression (GPR). In online phase our system, GPR gives prediction for location pose vector. helps therefore limits searching region leading reduce computational cost. An analysis on distribution k nearest points also presented which aims evaluate confidence extracting list trustable points. Furthermore, we present an extrapolation process this trust-list optimized trajectory. accuracy timing proposal was realized challenging BBIL dataset contains very noisy signals due fast moving object. experimental results show that exhibits much better than traditional or WkNN algorithms. RMSE optimal trajectory 1.78 m room dimension 10 $\times$ 25 m, competitive comparison other methods where initial known.