Chameleon: Survey-Free Updating of a Fingerprint Database for Indoor Localization

In fingerprint-based indoor localization, keeping fingerprints current is important for localization accuracy. Because access point (AP) signals can change over time or suddenly, the traditional approach is to survey the site regularly and frequently, which is laborious and costly. The authors propose Chameleon, a novel survey-free approach to localize users and maintain an updated fingerprint database despite fingerprint signal changes. In Chameleon, clients are both location "queriers" and unconscious "surveyors" (implicit crowdsourcing). Independent of and amenable to any fingerprint-based localization algorithm, Chameleon employs a clustering algorithm to filter out the altered AP signals to achieve high localization accuracy. Using the calculated location and the collected signals of the client, the fingerprint database can be continuously updated without manual intervention. Extensive experimental trials at The Hong Kong University of Science and Technology and at the Hong Kong International Airport confirm that Chameleon can adapt the radio map to the current signal environment and achieve low localization error despite signal changes.