Improved AdaBoost-based fingerprint algorithm for WiFi indoor localization

Indoor localization have received increasing attention for location-based severs in indoor environment. In this paper, we propose an indoor localization technique based on improved AdaBoost algorithm. The accuracy of AdaBoost depends on the weak hypothesis form all the weak learning, if there is noise in the fingerprint map, the performance of AdaBoost will decline. Because of the variability of indoor environment, the noise can not be avoided. So the improved AdaBoost is proposed to ignore the individual unfocused points to develop the localization accuracy. Experimental results indicate that the proposed algorithm achieves high localization accuracy.