A self-calibrating bidirectional indoor localization system

In this contribution an improved approach for indoor localization based on low cost Bluetooth Low Energy (BLE) systems is proposed. The measured Received Signal Strength Indicator (RSSI) values, which are available in Bluetooth systems anyway, are used for localization. It is well-known that localization based on RSSI measurements can only be achieved with a limited accuracy when applying these values directly to respective path loss models. Therefore additional signal processing is performed on the RSSI values and an auto-calibration approach is proposed, which takes inconsistencies with respect to environmental textures into account. The respective model parameters can be thus matched and updated in regular intervals. Further improvements are obtained by performing distance estimation in both directions, namely from transmitter to receiver and vice versa. Furthermore a sophisticated post processing of the RSSI values is performed. Compared to established RSSI-based localization methods based on low-power, low-cost systems, the localization accuracy has been significantly improved by applying the proposed methods. This is verified by measurements using an experimental setup.