Novel visible light positioning techniques : a thesis presented in partial fulfilment of the requirements of the degree of Doctor of Philosophy in Department of Mechanical and Electrical Engineering at Massey University, Albany, New Zealand

Abstract

Localization is the process of finding an object’s position within the space that it is situated in. Localization can be categorised into two types, indoors and outdoors. Outdoor localization is already a matured technology which mainly relies on well-known positioning satellite systems such as Global Positioning System (GPS) and GLObal NAvigation Satellite System (GLONASS). However, the indoor localization is still a growing area of research. Visible Light Positioning (VLP) has been getting the attention of researchers due to several advantageous factors. VLP is more accurate than many of the competing techniques. As Light Emitting Diode (LED) based luminaires have become an integral part of the indoor lighting systems in modern buildings and residences, such lighting infrastructure can be leveraged for localizing objects. The VLP systems are also suitable in places like hospitals and airports due to the fact that LED does not generate electromagnetic interference which can potentially affect the operation of many equipment used in those places. This doctoral research develops novel techniques and applications for VLP, and these are fully supported by experimental results and data analysis. Fingerprinting is a common positioning method used in VLP systems that employs Received Signal Strength (RSS) as the signal characteristics. Weighted K-Nearest Neighbour (WKNN) is one of the most popular algorithms for such localization systems. This thesis investigates the impact of distance metrics used to compute the weights of the WKNN algorithm on the localization accuracy of the VLP. Experimental results show that Squared Chord distance is the most robust and accurate metric and significantly outperforms the commonly used Euclidean distance metric. Robot navigation is one of the many potential applications of VLP. Recent literature shows a small number of works on robots being controlled by fusing location information acquired by VLP that uses rolling shutter effect camera as a receiver with other sensor data. In contrast, this thesis reports the experimental performance of a cartesian robot that was controlled solely by a VLP system using a cheap photodiode-based receiver. Two different methods (Direct Method and Spring Relaxation Method) were developed to leverage the VLP as an online navigation system to control the robot. The experiments consisted of the robot autonomously repeating various paths multiple times. The results show that both methods offer promising accuracy, with Direct Method and Spring Relaxation Method reaching the target positions of median / 90-percentile error of 27.16mm / 37.04mm, and 26.05mm / 47.48mm respectively. The operation of VLP is very much dependent on the line of sight (LOS) link between the luminaires and the receiver. Unfortunately, in a practical environment, luminaires are positioned to serve illumination needs. Therefore, enough luminaires may not be visible for the purpose of positioning the target. One way to compensate this would be to utilise an ultrasound system to eliminate the “blind spots” of the VLP system. The final part of this work consists of a study of the ultrasound based indoor localization. A bespoke system employing an ultrasonic array to transmit chirp signals and time of flight measurement for ranging was developed. The position of the receiver is estimated iteratively using the spring relaxation technique. The spring relaxation technique, which has not been used for ultrasonic localization in the literature, outperforms the widely adopted linear least square-based lateration technique. The experimental results show that the ultrasonic system can be a viable option for fusing with a VLP system.