Privacy-preserving Wi-Fi fingerprinting indoor localization

Abstract

A diversity of indoor localization techniques have become accurate and ready to use. A client first measures its location characteristics, and then calculates its location with the information provided by the localization server. However, this process may reveal the location information of the client or leak the area information stored on the server. This hinders the growth of indoor localization, but there are only a few solutions available in the literature. In this paper, we formulate an adversary model for fingerprint-based indoor localization techniques, and propose a cryptographic scheme to protect the privacy of both parties in a localization process. Our scheme utilizes the current Wi-Fi fingerprint based positioning techniques, including Gaussian radial basis functions and sigmoid kernels.