Improved Neighboring Reference Points Selection Method for Wi-Fi Based Indoor Localization

Abstract

Reference point (RP)-based methods such as K nearest neighbor (KNN) and k-means are frequently used for indoor location using received signal strength indicator fingerprints. However, these traditional RP-based algorithms (also termed clustering algorithms) do not use the information of geometric proximity between RP and test point for position determination. Meanwhile, the number of clusters needs to be predefined, which means an unsuitable number of clusters would lead to poor estimation accuracy. In this article, in order to eliminate incorrect neighboring RPs and to avoid selected RPs located only on one side of the test point, the geometric proximity between the neighboring RP and test point is analyzed in the online phase. The nearest neighboring RPs are selected based on their physical distances to the test point, instead of the widely used RPs' positions. The proposed algorithm was tested by the experiments conducted within an office building, and the results indicate that the proposed method significantly outperforms the traditional KNN, weighted K-nearest neighbor (WKNN), and test point irrelevant clustering algorithm.