Improving Wi-Fi Based Localization Using External Constraints

Wi-Fi based localization enables detection of users' position in indoor spaces by means of wireless networking infrastructure. The positive aspects of this solution include the reuse of already deployed systems and thus its reduced costs. On the negative side, Wi-Fi based localization is not particularly accurate, because the common operating conditions are far from the ideal ones. We propose to use external constraints for improving the accuracy of Wi-Fi based localization. A set of known schedules is used to restrict the estimated position of the user to a single room. The schedule for a given user is automatically selected from a set of possible ones by observing user's movements with coarse-grained resolution.