ANFIS-Based Wireless LAN Indoor Positioning Algorithm

This paper presents a wireless local area networks (WLAN) indoor positioning method based on adaptive neurofuzzy inference system (ANFIS), and improves the problem of decreased positioning accuracy caused by non-line-of-sight dominated transmission in commonly used indoor positioning systems effectively through the fingerprint method. First of all, the paper analyzes the design principles and mathematical model of ANFIS. Second, it solves the structure identification and parameter identification problems of ANFIS through subtractive clustering method and the BP network optimization algorithm. Finally, the paper makes system simulation based on the experimental data by applying the new method in WLAN indoor environment with the results of small system average error and good generalization ability.