An Accurate and Calibration-free Approach for RSS-based WiFi Localization

Abstract

We propose an accurate and calibration-free WiFi localization approach using received signal strength (RSS) to mitigate the impact of RSS variations caused by changing environment and heterogeneous hardware. In the online phase, we first divide a positioning area into G grid points and index each grid by a label. All the indexes of grid points form a label set. We collect labeled RSS fingerprints to construct an offline database at all grid points. Then, we first construct a supporting set (SS), which is a subset of the label set, selected by the similarity between the online RSS sample and offline database. So, SS is a latent space that likely includes the true label (location) of the user. We then derive an expectation maximization (EM) algorithm by incorporating the fingerprint quality into the estimation of the true label. Furthermore, we propose an optimal size selection (OSS) algorithm using bayesian information criterion (BIC) to adaptively determine the size of SS. Experimental results verify that the proposed approach performs significantly better than other existing methods.