Analysis of environmental factors for a WLAN location system using an estimated radio map

A WLAN location system based on an estimated radio map leads to cost saving for indoor location services. However, how to build an accurate radio map in complex indoor is a difficult task. Because such building structures as walls, ceiling, floor and pillars, which we call PAF hereafter (Principal Affected Factor), affect signal distribution by reflection, transmission and diffraction, these factors are processed to build an estimated radio map. The classical log-normal propagation model is modified by adding the PAF. In real-time location estimation phases, by matching captured RSSI with the above radio map, location result is obtained. Compared with the radio map without processing the PAF, it is shown in this paper that estimated errors on locations decrease in our several experimental sites by using our PAF based on calibration.