Abstract
In this paper, we consider a disaster scenario where a Micro Aerial Vehicle (MAV) is flying around the urban area and tries to localize wireless devices such as mobile phones. There is a high chance of those devices being in the vicinity of their human owners. Fast and simple approach to map the received signal strength to distance is the Received Signal Strength Indicator (RSSI). The more accurate mapping ensures higher localization accuracy. As a consequence, an accurate signal propagation model is required.
The Free Space model, ITU indoor and outdoor model, SUI model, Hata model, COST-231 Hata model and Log-distance model have been chosen to be investigated in this work. The goal was to determine whether analytically chosen models fit to our scenario, as well as develop a suitable model for outdoor-indoor scenario. A real-world experiment was carried out to collect RSS measurements. An MAV was placed outside of a building while mobile phones were located inside a building. A measure for the evaluation was a root mean square error (RMSE).
The main contribution of this paper is an adapted log-distance model for GSM which is suited for outdoor-indoor scenario with the RMSE value of 6.05m. The ITU indoor model represents the second best fit to our measured data with the RMSE value of 6.3m.