Indoor person localization system through RSSI Bluetooth fingerprinting

The growth of wireless and mobile communications technologies offers new possibilities for context driven information systems. Specifically, nowadays, mobile phones are equipped with several radio-frequency technologies, like Global System for Mobile Communications (GSM),WiFi or Bluetooth. In this way, the idea of using them to create a location system arises. Typical location algorithms can extract relevant information about Radio Frequency (RF) signals and estimate the position of the device. In this paper, we present a study of Bluetooth signal as source of information in one of these location systems. We analyze its capabilities and we create a set of algorithms to transform Bluetooth data in order to improve the location process. The system consists of N nodes connected to a cable network disposed in a scenario divided in cells, where each of them is composed of a thin client PC with a Bluetooth device and a directional antenna. It works by inquiring the Received Signal Strength Indicators (RSSI) values of all the visible Bluetooth devices. Employing a fingerprinting technique, it determines the most probable grid cell for each device.