Investigation Of Indoor Location Algorithms Based On RFID

Radio Frequency Identification (RFID), also named E-tag, is a generic term for the technology using radio waves to automatically identify individual items and collect associated information. Since traditional location techniques are not suitable for indoor environment, nor satisfying the accuracy requirements, RFID technique becomes an excellent candidate for indoor location technique because of its advantages such as non-contact and non-line-of-sight nature. Therefore, it is the aim of this thesis to investigate the location problems based on wireless RFID tags.In the design of RFID indoor location system, due to the existence of serious random interference in indoor radio environment, it is very important to have location algorithms with high accuracy. This thesis adopts empirical path loss model based on numerous measurements to obtain the mapping between signal strength and geometric path. In order to eliminate the measuring error due to different readers, Chan algorithm is used to obtain two dimensional position estimation of tracking tag. Considering spatial characteristic of data association, the nearest neighbor algorithm is investigated in detail. Because of the data association in time domain, this thesis takes into account the tracking tag's past positions to reduce the random influence on positioning accuracy. For three dimensional location problem, an enhanced nearest neighbor algorithm based on spatial data association in time domain is put forward. In order to increase the positioning accuracy, the ray tracing technique is introduced to study different indoor wireless transmission environments. It is shown that, compared to the empirical path loss model and path loss model in free space, the path loss model based on ray tracing is more applicable to indoor location system.