Indoor Location Sensing Using Multiple Wireless Standards

Abstract

Indoor location sensing systems are crucial for Internet of Things advancements, providing essential location-based information for large-scale applications and systems. The existing literature usually involves designing and testing systems using a single wireless standard. Therefore, it is still uncertain what is the best wireless standard for indoor location sensing and whether there are any benefits to implementing cooperative location sensing. Therefore, an indoor location sensing system was developed, utilizing the RSSI fingerprinting method with K-Nearest Neighbours algorithm. The system was used to evaluate the individual performance of Wi-Fi, Bluetooth and ZigBee for indoor location sensing as well as validate if cooperative location sensing is feasible. For testing purposes, the test sites were split into grids of 1.5m by 1.5m size and the system made predictions on the correct grid position of the user device. The indoor location sensing system developed has an accuracy of 1.5m for a detection range of 10m. In terms of prediction accuracy, the system managed to achieve a maximum accuracy of 80.56% in non-Line of Sight scenario and a maximum accuracy of 100.00% in Line of Sight scenario. The prediction accuracy performance of the system in predicting the correct grid position varies depending on the combination of wireless standards used, with the best accuracy result being 100.00% when using a combination of Wi-Fi, Bluetooth and ZigBee in a Line of Sight scenario.