Accurate source localization approach based on illuminating LEDs in indoor environment

Abstract

We propose a new positioning system based on the received signal strength (RSS) fingerprinting approach. The RSSs of LEDs are influenced by some ambient parameters, such as radiation pattern, incidence angle, attenuation coefficient, multipath fading and so on. And these factors degrade the performance of some existing range-based localization approaches significantly. To overcome this drawback, we prestore the RSSmeasurements of different grids by offline training phase. In the online localization phase, we compare the instantaneous RSS with the prestored RSS measurements using correlation matching technique. Our proposed approach can improve the accuracy of localization significantly without calibration process. The simulation results show the efficacy of our approach. Introduction White light emitting diodes (LEDs) have been widely used because of lighting efficiency, eco friendliness, and lifetime. The visible light communication (VLC) based on LEDs can be used as lighting system and communication system simultaneously. Indoor localization system based on visible light communication has many advantages including low cost, high data rate, and high accuracy. For such reasons, indoor positioning systems using VLC have recently gained popularity as effective alternatives [1]. Many approaches have been proposed for indoor localization using LEDs including TOA-based [2], TDOA-based [3], AOA-based [4], and mixed-based localization approaches [5]. All the listed approaches decrease the errors of localization to some extent. However, most of them need calibration of received power or some auxiliary devices in their localization process. Comparatively speaking, the localization technique based on RSS measurements of LEDs can be easily implemented and it has low cost with slightly low accuracy [6]. Some existing RSS-based approaches need calibrate the received power to improve the estimate accuracy of range [1]. The calibration process not only makes the localization system impractical in practice but also brings other errors to the localization results. Additionally, the RSSs of LEDs are influenced by some ambient parameters, such as radiation pattern, incidence angle, attenuation coefficient, multipath fading and so on. And these factors degrade the performance of some existing range-based localization approaches significantly. To avoid the calibration process, we present a RSS fingerprinting based localization approach to improve the accuracy of indoor localization system. The steps of our approach are as follows: firstly, we prestore the RSS measurements of different grids by offline training phase; sencondly, in the online localization phase, we compare the instantaneous RSS with the prestored RSS measurements using correlation matching technique. And the estimates of target nodes are given by searching the maximum value of correlation coefficient. Our proposed approach can improve the accuracy of localization significantly without calibration process. The simulation results show the efficacy of our approach. Joint International Mechanical, Electronic and Information Technology Conference (JIMET 2015) © 2015. The authors Published by Atlantis Press 971