Inquiry-Based Bluetooth Indoor Positioning via RSSI Probability Distributions

Abstract

This paper presents an inquiry-based Bluetooth indoor positioning solution via RSSI probability distributions. A practical system architecture is designed after the Bluetooth protocol and profiles are studied. Weibull function is applied for approximating the Bluetooth signal strength distribution in the data training phase. The Histogram Maximum Likelihood position estimation based on Bayesian theory is utilized in the location determination phase. The results show the possibility of indoor positioning through inquiring the Bluetooth-enabled handsets in range. Weibull distribution improves the performance of fingerprinting. The practicality of the system architecture is also proved by the outcome of a test campaign.