Fuzzy Model Based on Dynamic Weights of APs for Indoor Localization

Indoor localization technology based on the received signal strength (RSS) of wireless access point (AP) has become very popular in recent years. Considering that the Wi-Fi signal is unstable and uncertain, and the contribution of different APs to the localization are different, a fuzzy indoor localization algorithm based on dynamic weights of APs is proposed in this paper. The Multidimensional-Scaling Growing Clustering (MSGC) algorithm is employed to generate fuzzy rules flexibly without any prior knowledge. In order to maintain a compact and interpretable fuzzy rule base, the membership functions are merged according to the similarity of membership degree, and the consequent parameters are tuned by using the recursive linear-least-squares (RLS) algorithm. The gauss kernel function is adopted to weight different APs, and the iterative method is employed to assign weights for APs in different regions. The experimental results show that the proposed method could not only weaken the influence of unstable APs on the final results, but also reduce the update frequency of the fingerprint data.