RSS-based Indoor Localization Via Wireless Network

With the rapid development and popularization of wireless communication,indoor localization via wireless network plays a growing role in social basic supports,which is an information service technology.It is the core issue to be addressed in indoor localization to increase the positioning accuracy and improve the user's experience.This thesis investigates the fingerprint matching algorithm and RSS-based passive indoor localization system.The main jobs are as follows:1.Firstly,A brief introduction is made about traditional indoor positioning technologies.Afterwards,three advantages of RSS-based positioning technology are summarized,including performance,cost and universality.2.In order to positing the mobile terminal when it isn't connected to the wireless access point,a passive indoor localization system via wireless network is proposed.The principle and model are also put forward.In the sequel,combining the numbers and time,RSS data is sampled to be analyzed and experimented in indoor scenario,and an appropriate sampling scheme is stressed.A prototype clustering algorithm is introduced to filter the discrete points for RSS.3.As the RSS emerging a left-screw and double-peak phenomenon,probabilistic KNN algorithm is proposed for passive indoor localization system.Priori probabilities are weighted in the Euclidean distance calculattion.Then,the actual environmental tests and algorithm experiments are carried out in the offline and online phase.The probabilistic KNN algorithm is proved that it had a higher accuracy and also suitable for passive positioning system.4.A passive indoor localization system is designed and implemented,including RSS data flow and components.The RSS transmits data in real time between client and server side through a hardware and software platform.The positioning result is shown in a graphical user interface in the end.